Ryan Henderson demonstrates Fiscal AI's MCP plugin for automated financial modeling and insider tracking
Summary
Ryan Henderson of Fiscal AI explains how his MCP plugin connects AI models like Claude to audited financial data, reducing token usage and improving accuracy. He demonstrates building a driver-based three-statement model for Remitly in minutes, incorporating consensus estimates and qualitative context. The platform also automates earnings call summarization, research report generation from 10-Ks and investor letters, and insider transaction screening. Henderson highlights the tool's ability to identify unusual financial items, such as Alphabet's $8.2B SpaceX loss, and generate market heatmaps. He argues that AI allows analysts to offload grunt work and focus on critical thinking and conviction building.
Key takeaways
- 1 track all the KPIs that physical AI tracks, active customers, uh send volume per customer, uh, the take rate, the, and that's gonna be sort of your revenue drivers
Unidentified speaker (speaker_1) · 15:05
- 2 There's in this case, it even says distortions to keep in mind. Prime day timing. Prime day moved into Q2 for most countries, where uh it was Q3 last year. So it's going to affect comps.
Unidentified speaker (speaker_1) · 22:56
Episode sections
- 0:00 – 2:01
MCP Plugin Setup and Benefits
Ryan Henderson explains that the Fiscal AI MCP is a paid data connector plugin for AI platforms like Claude and ChatGPT. He notes it provides audited financial data, improves accuracy, reduces response times by avoiding web searches, and lowers token usage.
- 2:01 – 12:28
Platform Overview and Discount
The host mentions a discount for listeners and highlights the platform's ability to cross-reference quantitative numbers with qualitative text from filings and transcripts. He emphasizes the importance of knowing the data source for trust and verification.
- 12:28 – 22:06
Financial Modeling Workflow
Henderson demonstrates building a driver-based three-statement model for Remitly using a simple prompt. The model includes historical data, forecasts based on consensus estimates and context, and bear/base/bull cases, all generated within minutes.
- 22:06 – 26:01
Earnings Call Summarization
The speakers discuss using the tool to summarize earnings calls, extracting verdicts, guidance, and Q&A themes. Henderson notes the ability to link back to specific timestamps in the original transcript for verification and deeper listening.
- 26:01 – 31:05
Research Report Generation
Henderson shows a generated research report for Adyen, which synthesizes data from 10-Ks, fund letters, and slide decks to provide a business overview, competitive landscape, and moat analysis, leveraging qualitative context from thousands of investor letters.
- 31:05 – 42:38
Insider Trading Screeners
The final segment covers using the terminal to track insider transactions across the investable universe. Henderson demonstrates how users can create custom alerts for insider buys or sales, delivered via email or Slack, to monitor market activity.
- 42:38 – 47:47
AI-Generated Market Heatmaps
Ryan Henderson demonstrates using Claude and Fiscal AI to generate a heat map of S&P 500 stocks with drawdowns exceeding 40%. The tool automatically pulls all-time high drawdown data and visualizes it, highlighting the disparity between the index's all-time highs and the poor performance of many constituents.
- 47:47 – 50:55
Automated Insider Transaction Reports
The speakers review a generated report of insider buys and sales for the week of November 6-12. The system prioritizes transactions by value, highlighting significant purchases by executives at CVS and KMI, and large sales at Strike (STRK), demonstrating the tool's ability to aggregate and clean messy insider data.
- 50:55 – 56:04
Deep Dive into Google's Earnings and SpaceX Investment
Henderson analyzes Google's latest earnings, noting a revenue beat driven by cloud growth and maintained CapEx guidance. He highlights a specific non-operating charge of $8.2 billion attributed to an unrealized loss from Alphabet's investment in SpaceX, showing how the AI can quickly identify and explain unusual financial items.
- 56:04 – 1:02:22
Expanding Data Coverage and User Workflow
Henderson explains that the Fiscal AI MCP now includes global fundamentals, KPIs, stock price returns, and qualitative data like conference calls. He announces upcoming macro data coverage and encourages listeners to test the tool to offload grunt work and enhance critical thinking in their investment process.
Claims worth verifying
Fiscal AI has become the data layer behind more than 70 fintech platforms.
Unidentified speaker (speaker_0) ·5:38
How to verify: Check Fiscal AI's official website or press releases for client/partner lists; verify via company announcements. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Ryan Henderson has co-hosted Chit Chat Stocks since 2018 with Brett Schafer.
Unidentified speaker (speaker_0) ·5:58
How to verify: Check the Chit Chat Stocks podcast website or Apple Podcasts for host credits and episode dates. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
The MCP is a paid plan with monthly or annual options and a free week.
Unidentified speaker (speaker_1) ·0:54
How to verify: Visit the Fiscal AI pricing page to confirm subscription tiers and trial availability. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Remitly went public in 2021.
Unidentified speaker (UNKNOWN) ·14:06
How to verify: Check Nasdaq or SEC EDGAR for Remitly's IPO date and initial filing. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Primary sourceRemitly is a digital remittance provider tracking KPIs like active customers and take rate.
Unidentified speaker (speaker_1) ·15:05
How to verify: Review Remitly's investor relations page or recent 10-K/10-Q filings for business description and KPIs. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Primary sourceAmazon's Prime Day moved into Q2 for most countries, affecting year-over-year comparisons.
Unidentified speaker (speaker_1) ·22:56
How to verify: Check Amazon's Q2 earnings release or transcript for discussion of Prime Day timing and its impact on revenue. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Fiscal AI archives nearly 50,000 investor letters from around 8,000 global investors.
Unidentified speaker (speaker_1) ·28:41
How to verify: Check Fiscal AI's documentation or contact their support for current archive statistics. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Ryan Henderson calculated an 18% IRR for Remitly assuming an 18x exit multiple.
Unidentified speaker (speaker_1) ·16:51
How to verify: This is a specific calculation claim by the speaker; verify by reviewing the generated model if accessible or noting it as a personal example. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Primary sourceAdyen is compounding faster than at any point since 2023.
Unidentified speaker (speaker_1) ·26:47
How to verify: Check Adyen's recent quarterly earnings reports for revenue growth rates compared to 2023 periods. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Adyen trades at its lowest multiple since its 2018 IPO.
Unidentified speaker (speaker_1) ·26:47
How to verify: Check Adyen's stock price history and valuation multiples (P/E or EV/EBITDA) since its 2018 IPO date. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Primary sourceFiscal AI is the data layer behind more than 70 fintech platforms.
Unidentified speaker (UNKNOWN) ·42:52
How to verify: Fiscal AI website or press releases; verify via company blog or marketing materials. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
The S&P 500 had 134 stocks down 40% or more from all-time highs in October.
Unidentified speaker (UNKNOWN) ·42:52
How to verify: Fiscal AI platform or S&P 500 constituent data from Yahoo Finance/Google Finance for October 2026; verify via market data provider. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Google's revenue was $127.6 billion, beating Fiscal AI estimates by just under 1%.
Unidentified speaker (UNKNOWN) ·51:11
How to verify: Google Q3 2026 Earnings Report (SEC Form 10-K/Q); verify via SEC EDGAR or Google Investor Relations. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Primary sourceGoogle's Cloud business grew 103% year-over-year.
Unidentified speaker (UNKNOWN) ·51:11
How to verify: Google Q3 2026 Earnings Report; verify via SEC EDGAR or Google Investor Relations. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Primary sourceGoogle maintained its Fiscal Year 2026 CapEx guidance at roughly $200 billion.
Unidentified speaker (UNKNOWN) ·51:44
How to verify: Google Q3 2026 Earnings Transcript or Press Release; verify via SEC EDGAR. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Primary sourceGoogle raised its Cloud backlog guidance.
Unidentified speaker (UNKNOWN) ·51:44
How to verify: Google Q3 2026 Earnings Transcript; verify via SEC EDGAR. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Primary sourceAlphabet recorded an $8.2 billion unrealized loss from its investment in SpaceX.
Unidentified speaker (speaker_2) ·53:52
How to verify: Google Q3 2026 Earnings Report (Investment Income/Loss section); verify via SEC EDGAR. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Primary sourceThe Fiscal AI MCP was launched approximately one year ago with only US fundamentals.
Unidentified speaker (speaker_0) ·57:15
How to verify: Fiscal AI product launch announcements or blog posts from late 2026; verify via Fiscal AI website. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Fiscal AI will launch macro data in the next week from the recording date (Nov 18, 2026).
Unidentified speaker (speaker_2) ·57:35
How to verify: Fiscal AI product update announcements or blog posts in late November 2026; verify via Fiscal AI website. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.
Do your own research
- Verify the subscription cost and free trial terms for the Fiscal AI MCP plugin directly on the Fiscal AI website.
- Cross-reference the Remitly financial model's assumptions (active customers, take rate) with Remitly's latest 10-K and 10-Q filings.
- Check the actual insider trading activity for CVS and KMI during the week of November 6-12 via the SEC EDGAR database or a dedicated insider tracking service.
- Review Alphabet's latest earnings report to verify the $8.2 billion unrealized loss attributed to the SpaceX investment.
- Compare the AI-generated heat map of S&P 500 drawdowns with independent data from Bloomberg or Yahoo Finance to assess accuracy.
- Read a sample of the 50,000 archived investor letters to evaluate the quality and relevance of the qualitative data provided.
- Test the AI's earnings call summarization against a known transcript to check for hallucinations or missed key guidance points.
- Evaluate the cost savings of reduced token usage against the subscription fee for your specific AI usage patterns.
Full transcriptexpand
0:00Unidentified speaker (speaker_0): I shared with you that I was a bit hesitant at first because it's yet another thing to try. But I fell in love with it and it has changed my life. So I want to kick it off with you maybe showing us how do you include it in your AI chat experience.
0:15Unidentified speaker (speaker_1): Yeah. And I'll I'll preface this by saying I was kind of the same. I think most people are kind of in a similar boat About where it's like maybe you don't know what an MCP is or a data connector or it's not quite as easy to visualize as uh you know our our stock research platform. But it once like you said, once you see it, you can't unsee it and you you can really realize how powerful it is. Um so we'll we'll talk through some examples. But yeah, connecting to it is pretty straightforward. I think pretty much every AI platform, Chat ChatGPT, GPT, Claude, Codex, Copilot, they all have these connector ecosystems, kind of like an app marketplace. If you just go to their connector ecosystem, type in fiscal AI, it'll you'll find our connector. The MCP is its own plan. So it's a paid plan. Once you have set up an account and subscribe to the MCP, it's either monthly or annual, and there's a free week. too if you want to test out the data.
1:17Unidentified speaker (UNKNOWN): Then
1:17Unidentified speaker (speaker_1): it just connects automatically. So it'll be look up fiscal AI, little connect button in the corner, boom. It tells you all the data sets that you get. And it's basically giving it's plugging Claude into this massive audited financial database. So the data data's is going gonna to be more accurate, much more accurate. It will be quicker response times because it doesn't have to search the web, which is a huge value add if you're sitting there and waiting four or five hours for for a response. Sometimes the the projects can take that long. It it can be a huge difference. And then uh it reduces the token usage as well. Uh
1:57Unidentified speaker (UNKNOWN): yeah,
1:58Unidentified speaker (speaker_1): there's there a lot of benefits to having just that real audited database.
2:01Unidentified speaker (UNKNOWN): I'm
2:02Unidentified speaker (speaker_0): gonna mention three things that came to my mind. First of all, there's a discount available for Talking Billions listeners and the link will be included in the notes. So anybody that wants to try it, you get an additional discount. And if you happen to be listening around Black Friday, there will be one more discount on top of it from what I know. So you can explore it as you're taking a little break over the holiday season and get to know it, how it works. Very easy to set up. It you know, MCP, all those things sound very intimidating. You're adding a plugin. And anytime you want to research, you just select it, and we'll show you in a second how it works. But it's sources. I like what you highlighted. So it's faster speed because it comes from one source, reliable source, and it is a reliable source. It's all your data that you guys are collecting. And you know where it came from, which matters to me, and I think. listeners, when you're doing research, you want to know that you can trust what you're relying on. And there are ways to double-check what's coming through. I want to highlight, and it will come across throughout this conversation, that you have both the quantitative and the qualitative side of the experience, right? You have the numbers, but you also have the transcripts, you have the earnings releases, you have the text, you have obviously the filing. So any question you have, it cross-references both the numbers and the actual narrative of the story, which can be really
3:21powerful, and we'll talk more about it. The better questions you ask, the more value you can get get out of the experience. Welcome to Talking Billions. We talk about big ideas, big inspirations, big topics. We take on the hardest topic of all, money. How to make it, save it, keep it. But our conversations lead us to an even bigger question: what it means to live a rich life beyond money. My guests share their practices, principles, and evergreen wisdom. I'm your host. Bogomil Varanofsky, author, TEDx speaker, and investment advisor to wealth creators with patient capital and an infinite investment horizon. I work with families and individuals who aspire to grow wealth over a lifetime and generations through disciplined, thoughtful investments in durable quality businesses, while giving money, meaning. Join me on this quest to unearth and share the wisdom of the ages. Let me share with you the podcast program disclosure statement. Blue Infinitas Capital LSC is a registered investment advisor, and the opinions expressed by the firm's employees and
4:39podcast guests on the show are their own and do not reflect the opinions of Blue Infinitas Capital. All the statements and opinions expressed are based upon information considered reliable, although it should not be relied. upon as such. Any statements or opinions are subject to change without notice. The information presented is for educational purposes only and does not intend to make an offer or solicitation for the sale or purchase of any specific securities, investments, or investment strategies. Investments involve risk, and unless otherwise stated, are not guaranteed. The information expressed does not take into account your specific situation or objectives. and is not intended as recommendations appropriate for any individual. Listeners are encouraged to seek advice from a qualified tax, legal, or investment advisor to determine whether any information presented may be suitable for their specific situation. Past performance is not indicative of future performance. None of what you're about to hear is investment advice. My guest today is Ryan Henderson, head of research at Fiscal AI, the AI powered stock research platform, which has become the data layer behind more than 70 fintech platforms, and the co host of Chit Chat Stocks, one of the largest stock investing podcasts in the world, which he has hosted since 2018, with his former Washington State
5:58University football teammate Brett Schaefer, after previously running a real money. investment fund and now spending his days dissecting everything from one hundred beggars to whether a quarter of the SNP five hundred hiding in a forty percent drawdown spells opportunity or danger. How are you, Ryan? So good to see you.
6:17Unidentified speaker (speaker_1): I am doing well. I appreciate the intro and I appreciate the invitation. I'm a longtime listener, first time guest.
6:25Unidentified speaker (speaker_0): Super excited. You and I had many chats and many conversations. You've been showing me how Fiscal has been evolving and growing and all the new tools that you guys developed. First of all, it's a great platform to get access to historical data, live data, transcripts, everything about the companies. But now you have something even more exciting that I wanted to ask you more about. You and I had a deep conversation. about it a week or two ago. You convinced me to try it this summer. And I it's something I told you, once you see it, you cannot unsee it. And we're talking about the AI plugin, which allows users to use the platforms they use, you know, chat or cloud or whatever you guys use, plus have access to fiscal AI data. So you have a reliable source of information. I've had many guests on the show talking about fundamental research. I'm a big fan of fundamental research. It's my daily life managing money for people. I had Robert Hackstrom, Chris Mayer, Paul Johnson, Chris Beck. Half of them are teaching, investing to others. I had Essa Nissani, who you might know, who is also very curious about AI, and
7:37the topic keeps coming back. The tools keep on improving, but you guys really stand out in that space. I do want to highlight for the listeners that Preethan Dennis was on the show and he got a gave an incredible overview of Fiscal, the platform itself. So for anybody listening, look up that video. You're gonna see a lot. But today we're gonna just talk about the plugin, how it works. I shared with you that I was a bit hesitant at first because it's yet another thing to try, but I fell in love with it and it has changed my life. So I want to kick it off with you maybe showing us how do you set it up. It's a lot easier than people think. How do you include it in your AI chat experience?
8:20Unidentified speaker (speaker_1): Yeah, and I I'll preface this by saying I was kind of the same I think most people are kind of in a similar boat where it's like maybe you don't know what an MCP is or a data connector or it's not quite as easy to visualize as uh you know our our stock research platform. But it once like you said, once you see it, you can't unsee it and you you can really realize how powerful it is. Um so we'll we'll talk through some examples. But yeah, connecting to it is pretty straightforward. I don't know if I can um I don't know if screen share necessarily makes sense for it, but basically for all I think pretty much every AI platform, ChatGPT, Claude, Codex, Copilot, they all have these connector ecosystems, kind of like an app marketplace. If you just go to their connector ecosystem, type in fiscal AI, it'll you'll find our connector. The MCP is its own plan. So so it's a paid plan. Um once you uh have set up an account and subscribe to the MCP, it it's either monthly or annual, and there's a free week too if you want to test out the data.
9:35Um, then it just connects automatically. So it'll be Boom, it tells you all the data sets that you get. And it's basically giving it's plugging Claude into this massive audited financial database. So
9:55Unidentified speaker (UNKNOWN): the
9:55Unidentified speaker (speaker_1): data's gonna be more accurate, much more accurate.
9:59Unidentified speaker (UNKNOWN): Uh uh
10:00Unidentified speaker (speaker_1): it will be quicker response times because it's it doesn't have to search the web, which is a huge value add if you're sitting there, you know, waiting four or five hours for for response. Sometimes the the projects can take that long. Um it it can be a huge difference. And then uh it reduces the token usage as well in instead of having to go through the open web. You're not consuming quite as much tokens. So uh
10:26Unidentified speaker (UNKNOWN): yeah.
10:26Unidentified speaker (speaker_1): there there's a lot of benefits to having just that real audited database. Um and and yeah, it's pretty s I think it's pretty straightforward. There's also some videos online of how to connect um that I can send you Bogomil and maybe we can link to them or something. But uh yeah, is that to kind of cover it?
10:44Unidentified speaker (UNKNOWN): It
10:45Unidentified speaker (speaker_0): covers it. I'm gonna mention three things that came to my mind. First of all, there's a discount available for talking billions. listeners and the link will be included in the notes. So anybody that wants to try it, you get an additional discount. And if you happen to be listening this around Black Friday, there will be one more discount on top of it from what I know. So you can explore it as you're taking a little break over the holiday season and get to know it, how it works. Very easy to set up. It you know MCP, all those things sound very intimidating. You're adding a plugin. And anytime you want to research, you just select it, and we'll show you in a second how it works. But it's sources. I like what you highlighted. So it's faster speed because it comes from one source, reliable source, and it is a reliable source. It's all your data that you guys are collecting. And you know where it came from, which matters to me, and I think listeners, when you're doing research, you want to know that you can trust what you rely you're on. relying on. You don't want to
11:40Unidentified speaker (UNKNOWN): and
11:40Unidentified speaker (speaker_0): there are ways to double check what's coming through because there are links you're gonna probably show us today, how you can see where the data came from and see the whole data set if you want to and the original source and everything else. I want to highlight and it will come across throughout this conversation that you have but both the quantitative and the qualitative side of the experience, right? You have the numbers, but you also have the transcripts. you have the earnings releases, you have the text, you have obviously the filing. So any question you have, it cross-references both the numbers and the actual narrative of the story, which can be really powerful. And we'll talk more about it. The better questions you ask, the more value you can get out of the experience. You want to start with the financial modeling? I know you have something ready for us. to look at.
12:28Unidentified speaker (speaker_1): Sure. Let me uh share my screen here for anyone watching and then I'll kind of talk through it as well. So this is probably one of the most common use cases that I've been personally leveraging the MCP for. And again, MCP for the average investor like myself, kind of an intimidating acronym. or at least a new acronym, but it's just an AI data connection. So it's a plug-in to your AI for a database. Uh and with the MCP, when you uh when you connect, whether you're in Claude or ChatGPT, it comes with pre-installed skills. So in and a skill is just a pre-built prompt. So when you uh once you've connected, you could say build me a financial model for Remitley. That this is a company I own. So it was one of the first ones I I modeled out. And it'll use our base pre-built modeling prompt. So it kind of helps people you know uh visualize some of the use cases.
13:35But it the model is very thorough. Uh like I think peop I really encourage people to test it out and just see what it can provide. It's gonna have your your cover page. The drivers is where investors should probably spend the most time. And and it what it's going to do is it's going to bring a bunch of historical financial data. So it depends on the company how long they've been public. In this case, you've got five years. of trailing historical data,
14:06Unidentified speaker (UNKNOWN): uh,
14:06Unidentified speaker (speaker_1): just because room at least I think went public in 2021. And then it's got its own forecasts that it's generated. And the forecasts are going to use a combination of consensus estimates, uh statistical regressions, and context. The third part here is really important. It's used using context from conference calls. It's using context from fund letters, from filings, anything it can find to be relevant for that specific estimate, it's going to utilize and it's going to build you a base case, a bear case, and a bull case. And it's uh going gonna to default to the base case. And this is pulling in in this case, Remittly is a digital, mostly mobile uh remittance provider. So primarily people that have moved to the US, they send money back abroad to families back home or friends. Um and so it's gonna track all the KPIs that physical AI tracks, active customers, uh send volume per customer, uh, the take rate, the, and that's gonna be sort of your revenue drivers, then it's gonna go through all the expense lines and kind of your standard model. Uh and all these are very easy to edit.
15:25So, you know, if you think, oh, I think active customers are gonna grow at instead of 13%, 12%, 11%, 10, I think it's gonna be higher, you know, 15, 14, whatever. Very easy to edit, and you could just do that there, and then it's gonna just fill through the rest of the model. It's got all the uh income statement, balance sheet, cash flow, uh historical data. So you can easily cross-reference if you want to see like, you know, you're guessing what active customer growth is gonna be. Well, let's check what it was for the last four or five years. And then you've got sort of your extractions. here or your your forward projections in terms of return. So you've got a DCF, and then this is kind of the one I use. You just, and I know this is lazy, some investors maybe frown
16:14Unidentified speaker (UNKNOWN): upon
16:14Unidentified speaker (speaker_1): this, but you slap a multiple on it. Uh, and what would it look like, you know, three, four or five years out? So uh you can just adjust your expected multiple and it's gonna going to give you uh if we scroll. down here, expected returns, if I can find it. Um
16:33Unidentified speaker (UNKNOWN): I'm
16:33Unidentified speaker (speaker_1): looking through I'll find it here in a second. But yeah, so base case for me, and kind of talking my book here, but uh if I assume they have a earnings multiple of 18 times uh uh on exit, it'll be an 18% IRR. Again, this is kind of just one of my companies, but you can do this for literally any company. It is that simple to say, Claude, build me a financial model for ADN, for Amazon, for Airbnb. It it's really powerful. I have built just a Google Drive full of models, and each one of these takes about 15 to 20. minutes, which might sound like a lot, but when you think about compiling all this data yourself, uh it's pretty
17:22Unidentified speaker (UNKNOWN): small in the grand scheme weeks.
17:25Unidentified speaker (speaker_1): Yeah.
17:27Unidentified speaker (speaker_0): Can you can you show the prompt that you use to generate something like that? Just to give us an idea. Would you be able to share a tab that has the private prompt?
17:41Unidentified speaker (speaker_1): Yes.
17:44Unidentified speaker (UNKNOWN): We don't have to
17:44Unidentified speaker (speaker_0): run it just to give an idea for the benefit of the audience what it looks like.
17:49Unidentified speaker (speaker_1): Yeah, let me pull up Claude here.
17:52Unidentified speaker (UNKNOWN): So for listeners
17:53Unidentified speaker (speaker_0): that are not watching the video, when you get home, watch the video. But if you're listening on the road, we're looking at a very elaborate financial model with lots of tabs. with all the details. And it's all generated based on fiscal AI data. And right now Ryan is showing us his clawed. And we're going to go over the prompt. You may want to read out the prompt and then
18:20Unidentified speaker (UNKNOWN): just
18:20Unidentified speaker (speaker_0): show us the trick, how you involved Fiscal in this prompt, because that's something that's becomes intuitive. after a while but at first it's like how does fiscal join the conversation
18:31Unidentified speaker (speaker_1): yeah so my model was or my prompt was literally this simple build me a financial model for remitly it knows to default to fiscal AI as the data connector so if you've already connected a database it's going to default to it because it knows it's pulling uh authentic audited data so mine was literally this simple, but I think one of the beautiful things about using AI for investment research, everyone wants to optimize for like, you know, here's the best way to prompt and all this stuff. You can literally word dump whatever you want, and it's going to, you know, it's going to be able to provide you a valuable response. Or the thing I often do is ask it what would be a good project. prompt if this is the outcome I want? What would be a good prompt for a financial model? And it'll give you kind of the full detailed prompt. The again, this is a very base prompt, build me a financial model for Remitly. But if you said, assume customers are going to grow at 20% a year, you
19:33can add all your inputs on and you can do it as much as you know, you can be as thorough or as simple as you want. Um, and then you you probably just want to check that fiscal AI is connected, which in
19:45Unidentified speaker (UNKNOWN): this
19:45Unidentified speaker (speaker_1): case it is. Um, and then you'll see, like as soon as I run this, it's going to automatically pull up a fiscal AI logo and say, checking from yeah, okay, it
19:57Unidentified speaker (UNKNOWN): says
19:57Unidentified speaker (speaker_1): I'll build a driver-based three three-statement statement model from fiscal AI data. So you can see how it's kind of thinking.
20:04Unidentified speaker (speaker_0): So you don't have to add the word fiscal even in the prompt just because the plugin is active, it's already pulling fiscal data.
20:14Unidentified speaker (speaker_1): Yeah, you could be uh you could add it if you want, but yeah, it's gonna default to it. And if you're watching it, like you could just run it in the background, but if you're watching it, you'll see here's what it's thinking through, here's where it's pulling. this data, and it'll kind of talk through its its thought process.
20:32Unidentified speaker (speaker_0): So I almost think of it as a library, right? So it's when you ask the same question and you don't have fiscal plugin available or open or active, it's trying to build it from all kinds of data out there, whatever it finds. Here it goes to fiscal, and at least you know it's the reliable data that you guys collect and that you can see in the platform as you're experiencing it anyway. But here it's being used to produce an outcome, which in this case it's a fairly elaborate financial model. And I've been investing for 20 years. I know what it takes to build a model from scratch. And uh yeah, I have a lot of respect for the tool.
21:12Unidentified speaker (UNKNOWN): Yeah,
21:12Unidentified speaker (speaker_1): and I think one of the beautiful things is our our data, physically eyes data, is it all it's all sourced. So
21:20Unidentified speaker (UNKNOWN): uh
21:21Unidentified speaker (speaker_1): if you're on the terminal, for example, you can see you can just click on any metric and it'll pull you to that exact page in the filing that that number came from. You can add that as well, uh, in your models. So it's gonna have all the all the filings attached to the metrics right there, so you don't have to go hunt down where did this number come from to verify? You can just click on the source.
21:46Unidentified speaker (speaker_0): Ryan, we all spend a lot of time following our companies and we look at earnings. There's a lot that that obviously is available in the release, a lot of numbers, but I feel like the magic happens if you can know and follow what's been said and how it's been said. You guys have a tool that allows you to summarize earnings transcripts. Just get to the bottom of it and see what was actually said. Do you want to share with us what it that looks like?
22:13Unidentified speaker (speaker_1): Yeah, I exported the uh result here and it kind of came in a little f weirdly formatted At it just because it didn't port to Google Docs correctly. But if you're reading it within Claude, you'll
22:27Unidentified speaker (UNKNOWN): in
22:28Unidentified speaker (speaker_1): this case I saved uh a skill, which I just said I just went through the process that we just described where I said, if I were an analyst and I wanted a really thorough earnings call recap, what would be a good prompt? Claude sends me back a prompt. I save it. I say every time I ask for an earnings recap, use that prompt. So I did that for Amazon. It's gonna going give to you kind of an overall verdict. It'll say what the results were compared to expectations. There's in this case, it even says distortions to keep in mind. Prime day timing. Prime day moved into Q2 for most countries, where uh it was Q3 last year. So it's going to affect comps. Um, it's going gonna to say what the guidance was for uh the upcoming course. quarter and the upcoming year. And then it's going to you can go through QA themes. So if you want, you could say and in this case they're actually all timestamped too. I brought it into Google Docs so you can't click on them, but you can click right into it and see the exact commentary if you want, because this is polling from our earnings
23:35transcripts uh in the database. And then this case it kind of just gave a summarization for the Q and A. But if you wanna be really specific and you wanna say, give me the exact questions, um, you can make it as thorough or or as simple as you like. And it's a uh for me,
23:54Unidentified speaker (UNKNOWN): you know,
23:55Unidentified speaker (speaker_1): for your top holdings, y You
23:57Unidentified speaker (UNKNOWN): you're
23:57Unidentified speaker (speaker_1): you're probably going to read the full call, maybe. Um, I guess depends how many companies you own. But uh if it's one where it's one you follow and maybe you don't have time to read the entire call, but you want just to get caught up up to speed on what's happening, what did analysts say, what did management say that maybe changed how the stock reacted, this is a really useful tool and it it's something I do for most of my watch list, basically.
24:28Unidentified speaker (speaker_0): You know, I I shared it on the show and I might have told you before that I want to hear how things were said. And this tool, which is for those that are just listening and cannot see it until they get home and watch the video, it's a very detailed report that writes up what was said during the call. Now, what I like about it is that I can go back and ask specific questions. What did they say about an item that I care about or I've been watching for a while? And then I can go back and listen to write time in an hour-long recording and then hear it. It's something that I like, and I've met people that appreciate it too, maybe listeners as well, that you can hear it in their voice, in the management's voice. How did they actually say it? And in the world where we all have the same information. the same details, the same everything, your ear is your own ear. I feel like it can give you an insight. Now listening to every hour of every single call of every single stock you want to own or you own, there are only that many hours in the day and I don't have more hours than anybody else, right? So if I can zoom in and listen to the three minutes when they're talking about something I really, really want to hear about
25:39I it adds a huge value to me. So that's how I look at it. And I appreciate you showing it to us. Ryan, uh, I'm curious about research reports. It's also a way to produce a report that can get us started or maybe give us a refresher on an idea we've held for a while. Do you want to show us how that works?
26:01Unidentified speaker (speaker_1): Sure, thanks. So yeah, I actually ran one. this morning on Ad Yen. Again, the formatting might look a little wonky here just since I imported it to my my Google Docs, but this is another one of our pre-built prompts. So one of our uh default skills that comes with the MCP. And we've built a prompt that we think would generate a good comprehensive overview of a business. And this for me you know, if you if you're an expert on a business and you've followed it for a long time, maybe you're not getting as much value out of this. But if you're starting your research from scratch and you just want to go from zero to one, this is really helpful for getting up to speed on a business. And it's going to give you a summary of sort of where we're at today. So in this case, I ran this for Adion ADN and says Adion is compounding faster than. at any point since 2023, yet the stock trades at its lowest multiple since its 2018 IPO. Pretty good start, I guess, for enticing potential investors, maybe.
27:09History and business model. And it's pulling this from 10Ks, it's pulling this from fund letters, it's pulling it from uh slide decks or investor presentations. It's all that you mentioned it at the top of the show. The having the raw financial data is great for like data retrieval, but if it i you need context, like it you need a narrative and numbers. So th by us being able to feed qualitative data essentially, whether that's calls, filings, fund letters, stuff with opinions in it, um, it's able to extract more value uh for the the end customer. Um it's gonna going go to through the competitive landscape. In this case it's Ad ADN, EN, PayPal block, it's even mentioning Stripe as a private company, although it can't give too much detail on Stripe since the data's not reported as regularly.
28:10Unidentified speaker (UNKNOWN): And
28:11Unidentified speaker (speaker_1): then it's going to go through Moat. This is where I think it uh leans on some of the fund letters. So it'll look through what have people identified as kind of competitive advantages.
28:22Unidentified speaker (UNKNOWN): Uh
28:24Unidentified speaker (speaker_1): you know, they they maybe don't talk about it as much in conference calls. or in in 10Ks, but uh on fund letters, that's, you know, very common. So they can pull from that. And and the fund letters, I'll mention, I think it's around 8,000 investors globally, uh, that
28:40Unidentified speaker (UNKNOWN): it's
28:41Unidentified speaker (speaker_1): they're pulling from. There's I think almost 50,000 total letters Little that are are in the archive now and it continues to grow every time someone writes something. Um and then it'll give you kind of financial updates, profitability KPIs as well. It's just a really quick way to get up to speed on a new business.
29:09Unidentified speaker (speaker_0): Report with your own pieces, the way you like to write it up. It almost to me, I mean, it does look like a really knowledgeable analyst that very quickly can write up a whole company. Is it everything you want to know? Probably not, but it's an incredible start. And then once you start interacting with it, you start to tweak it and you might have an in-depth report for a company that just talks about the mode, or just talks about the competition, or just talks about KPIs. And you can go really deep and do a cross-reference across peers, time, what's been said, how the conversation evolved. You pointed it out a couple of times, but I want to highlight that the qualitative side, the investors that you guys follow, and the letters that for many years we could all access through fiscal. Now this tool taps into that resource. So you not only have the story told by the company, which I think it's a great start, but you also have a wide range of opinions. So you almost have in your back pocket not just the the analysts that work for you, but also everybody that said and shared anything, you know, sensible, thoughtful about those companies out there. You may agree with them, you may de d disagree with them, but it definitely expands your perspective. And I think that adds a huge
30:29value. I always ask people to poke holes in whatever I said. Like whenever I give a talk or anything, I say, you know, take it for what it is, one person's work, poke holes in it. But with this tool you can really poke holes in what you think. And if you still like the stock, good for you. Buy a few shares. By the way, any stocks that you see on the screen today are no recommendations. So do your own research They're just case studies, you know, for the two of us to have a little bit of fun. Uh Ryan, building applications, screeners, there's so much more that you can do. Do you want to show us one more um case study or use case?
31:05Unidentified speaker (speaker_1): Yeah, I think maybe the best way to show this is actually just within within the fiscal fiscally AI I terminal, because it gives you a running list, but I'll talk through kind of what you can do here. So one of the things we track is insider transactions. So across the the whole universe of equities, we track who's buying, who's selling. We're basically looking through every form for and whatever the other forms are where they d disclose. Um and we're distinguishing between open market transactions, options vesting, that kind of thing. And one prompt I run that I think can be valuable is uh show me what insiders from across the investable universe bought stock
31:53Unidentified speaker (UNKNOWN): in the
31:53Unidentified speaker (speaker_1): open market in the last week. And you can just, and there's a lot of different delivery methods. I think that's something that maybe intimidates people. It's like, well, if it gives me a markdown file, what am I going to do with that? But you could say, send me an email every Monday morning with all the companies that had executives buy stock in the open market last week. And it's it's like you're you're getting a report from an analyst or you're getting like a curated news feed for whatever you like. And um I'm using insider trades in in this case, and and you can scroll. through and you can see. Okay, I have no idea what this company does, but uh AlphaTech Holdings had an independent director by
32:36Unidentified speaker (UNKNOWN): I
32:37Unidentified speaker (speaker_1): don't even know what the value is on it, but 35,000 shares or whatever. You can also customize it for insider sales. You can uh we also have politician transactions if you want to see what what politicians are. buying or selling stock, you can ask for that as well. All these are in that data feed. So it's there's a lot of different ways that you can do it, whether you want it sent to a Slack channel, whether you want it sent to your your email inbox, or um you just want a file uh drafted every week. It's it's very intuitive and and very customizable.
33:18Unidentified speaker (speaker_0): And it's also through Claude or Chad that you can interact with that data, right?
33:23Unidentified speaker (speaker_1): Correct. Yeah. I I I'm using the terminal in this case, but sort of the the back end powering a feature like this is essentially what we just talked about. Show me like all the insider transactions and it's just gonna going to give you produce a running list. So
33:37Unidentified speaker (speaker_0): we're looking at Claude and Ryan is showing. us how we can basically source the same kind of information but with our prompts.
33:48Unidentified speaker (UNKNOWN): And
33:48Unidentified speaker (speaker_1): you'll see it pull up the little fiscal logo if you're ever sure, like you know you want to double check
33:53Unidentified speaker (UNKNOWN): that
33:53Unidentified speaker (speaker_1): it's pulling.
33:54Unidentified speaker (speaker_0): As we're
33:55Unidentified speaker (UNKNOWN): looking
33:55Unidentified speaker (speaker_0): at it, you know, I kind of want to highlight because the more I think about it and I have so many guests to bring it up that that talk about the different uses of the new tools that we have. The data is out there, but I think the overwhelming thing is that there's more and more data. And that these tools allow us to interact with this data in a way that I think I can't imagine doing it 20 years ago when I started. So you had to choose and pick which stock you're going to gonna give that kind of attention. And now you can really double check your convictions, your assumptions. assumptions very quickly in various ways. And you know, the insider transactions, I I have friends that follow that much more closely than maybe I do at times, but it's something that was always tricky to digest. How do you get a report that's useful, that's you know, current? And with these tools, you can get the data, you can look at it, and you can decide if it's helpful. if it's an additional vote on of you know confidence. Or if it's actually undermining what you thought is happening and the insiders are leaving the ship and maybe you should be looking for a live vest.
35:03Unidentified speaker (UNKNOWN): Yeah,
35:03Unidentified speaker (speaker_1): that that's a good example too, is there's probably a lot of a lot of noise in insider transactions because you know the grandson of some founder that inherited a bunch of money is probably going to sell stuff stock pretty regularly or whatever. Um, but you can customize it, say, only show me insider sales where there's more than a million dollars or you know, whatever. Um and like you said, all this most of this data is publicly available. You could go to a company's investor relations site, you could go to the recent SEC filings, click through every file. form for yourself. Um, but to to get it aggregated, cleaned, standardized, extracted all the relevant context around the transaction. So, you know, was it an open market sale? Was it a derivatives vesting? How much was it? What was the average price? What's the role of this executive? All this stuff. Having it pulled into one standardized feed is pretty valuable.
36:08Unidentified speaker (speaker_0): It's a very powerful tool. And I think the more I interact with it, I ask more questions before I would have to think about who is actually going to do the work, right? So even if you have analysts working for you, you realize what's the highest priority for them? And some of the questions would not be even asked because there's just limited manpower to answer those questions. But in this case, you can continue to ask questions. And if anything, I think the quality of questions continues to improve. And I was under the impression initially that eventually we will all be getting the same answers. But then I realized we're not asking the same questions. And you know, just going back in time, I remember being in meetings with companies, and there were so many bright people in the room asking really good questions. And then I realized we come to the meeting, any meeting with a CEO, CFO, the leadership of a company, with our own experience, biases, circle of competence, and we will walk away with very different answers. And I remember walking out of the
37:13room and you could grab coffee with people. And among any random three, you would have somebody that's convinced that they should sell shares as soon as they get back to the office. You would see somebody who is convinced they should triple their holding. And you would have somebody else who would say, I'm just comfortable holding what I have. So AI tools don't give you the same answer if you ask better and better questions. They will give you the kind of answer that we will help you on your own journey. And the beauty of the market is that we can all disagree. And somehow, in Ben Graham's words, you know, the weighing machine continues to work.
37:51Unidentified speaker (speaker_1): Yeah, exactly. I I genuinely believe that there has never been a better time to be uh a stock research analyst. um or an analyst uh
38:06Unidentified speaker (UNKNOWN): I
38:06Unidentified speaker (speaker_1): guess of any investment securities because the grunt work is kind of removed. The the stuff that took you two weeks building that really thorough model takes 15 minutes now, and you can spend your time getting a better understanding from management, getting a better understanding of where is the business really going, what what are the opportunities, like the the valuable stuff you can spend more time on and you can offload the busy work essentially to AI. Um and then you can add your own inputs. Um yeah, it it really is kind of a kind of a wonderful world wonderful world for. equities analysts.
38:53Unidentified speaker (speaker_0): And it's a peculiar time because you and I know that passive investing is such a big portion of ownership of companies and activity in the markets. And, you know, I have nothing against it. I think it's a great way for people to participate in the ownership of wonderful businesses, just owning a what I would call a Noah's arc of them, which means, you know one of everything, which you know is a way to go about it. But if you do have the time, the attention, or if you're doing it professionally, you have better tools than ever to do th really thorough work and find out everything you possibly want to know and get the comfort around an idea to buy, sell, hold, whatever it is. And in my mind you know, the end goal is you buy something, you hold it for a period of time, and maybe you never sell it. In between, for me, it's all about conviction. What gives me conviction to continue to hold it? I've had enough guests on the show that tell me that they bought the right stock at the right time at the right
39:56price, but then on the during the recording or after the recording, they share with me, I I didn't get to hold. it long enough. And I think it happens to all of us. I think that the our investing lives are full of regrets of that kind where you had something that just continued to do better and better and you no longer hold it. So if you have tools that can nourish, nurture that conviction that allows you to hold something for a long period of time,
40:22Unidentified speaker (UNKNOWN): you
40:22Unidentified speaker (speaker_0): know, more more ma power to you. And you might not have an analyst working for you that will do the kind of work that creates
40:29Unidentified speaker (UNKNOWN): you
40:30Unidentified speaker (speaker_0): know Friday evening doubts about your holding, but you sit down with fiscal, you run a couple of questions and you realize, you know what? On Monday, I'm not gonna sell it. You know, I convince myself that I should hold it a bit longer. I think the conviction is that highest value added that we can have in the investor's life. And that's that's how I look at it.
40:52Unidentified speaker (speaker_1): Yeah, it is uh it it's an uncomfortable feeling. Well, it's an uncomfortable feeling having a stock that you really like go up a lot. Not not a feeling I'm super familiar with, but uh I imagine having the valuation rise a bunch, it puts you in a
41:10Unidentified speaker (UNKNOWN): sort
41:11Unidentified speaker (speaker_1): of an awkward position. And Having someone you can go back and forth with. with
41:15Unidentified speaker (speaker_0): and
41:16Unidentified speaker (UNKNOWN): having
41:16Unidentified speaker (speaker_1): data you can pull from, whether that's, you know, eight thousand other investors that you can that have produced written work for a company or whatever, it's kind of reassuring and uh helps you kind of hold that conviction.
41:32Unidentified speaker (speaker_0): Brightest, I'll share a secret for those that are still with us here. The brightest minds out there. Everybody you can think of. You know, I had Chris Davis on the show who's a Berkshire director and investor in his own name. Buffett himself had stories about it too. Everybody has a story of a stock they owned, and they they remember selling it and regretting it. He shared it here and there, and he did the math. then He wishes he never sold it. So there everybody has a story. Chris Davis talks about Costco. There are other stories in between. So what is it that you can do to maintain that conviction, even when the valuations are higher than something you're you know, comfortable with? And I feel like that kind of in-depth fundamental research can restore, Store recharge that conviction in moments of doubt. And the moments of doubt will happen again and again. But these are the tools that can help us. Ryan, as we're waiting for this, do you want to show us graphics? I know that you have some examples that you wanted
42:35to share with this audience as well.
42:38Unidentified speaker (speaker_1): Yeah, sure thing. And maybe we can jump back to that. I think the screen sharing on it might have been slowing it down, anyways. Um let's see if I can gonna pull this one up. This is one I built recently um that got a lot of engagement. Now some of the state is
42:52Unidentified speaker (UNKNOWN): out of date because I
42:53Unidentified speaker (speaker_0): like those colors. Yeah.
42:54Unidentified speaker (speaker_1): Yeah I I built it in October, but it was basically uh uh I can go double check the prompt, but it was basically show me the drawdowns, uh
43:05Unidentified speaker (UNKNOWN): show
43:05Unidentified speaker (speaker_1): me all the stocks in the SP 500 that are in more than a 40% drawdown. And I do a lot of work for fiscal where it's sharing content. So uh posting charts on Twitter or you know, uh putting them in a newsletter or sharing them on LinkedIn, whatever it is. And so I'm constantly building graphics. And I found that Claude can be an incre Claude, I guess, plus the fiscal MCP, can be so powerful for sharing investing graphics. This one automatically builds a heat map, pulls the total drawdown. from all-time highs. Now, granted, some of these companies, I think AIG might have set their high in 2001. So it's it's maybe not exactly indicative of recent recent performance, but um yeah, it's it's a powerful little visual. I I did one yesterday, ran one that was pull the top 50 fifty software. holdings from I I think it was the iShares uh software ETF, put them on a scatter plot comparing next 12 month revenue expectations versus uh the the forward e bit multiple. And within five ten minutes, it's pulled all the all the underlying s uh
44:25data that's required from there, whether that's forward eBIT multiple, forward revenue expectations. Again, both of those are being pulled from the fiscal data. Um, all the top holdings. We we just added fund and ETF holdings. Uh pulling all that and putting together a scatter plot, and it it makes it a lot prettier than me going into Canva and going one by one and changing the red, you know, changing the red by the slightest. hue. So it's like it saves so much time and it really does generate some powerful graphics. Um
44:57Unidentified speaker (UNKNOWN): yeah,
44:58Unidentified speaker (speaker_1): this this was one example, but I could scroll through Twitter and show you the hundreds.
45:08Unidentified speaker (speaker_0): Well, for the benefit of those that are listening, we're looking at a
45:13Unidentified speaker (UNKNOWN): at
45:13Unidentified speaker (speaker_0): a slide that has the 134 S P five hundred stocks that are down forty percent or more and it's by different shade of red from very intense red all the way to very bright red. But you see the tickers and the drawdown, which gives you an incredible visual, especially, you know, helpful when you're thinking about the market. Is it is it yeah, we had moments when this market was uh close to an all-time high or recent high and you feel like a lot of stocks are actually down. And it's different to say, oh, it's that many stocks, but to actually look at a picture, an image, and be able to share it. And I've seen this shared around, and I'm glad you brought it up. And for those that get home and watch the video, you can see it too. But a picture, sometimes you show the picture and you don't have to say anything more. You know, I think you can just put the slide up in a conference. room with people and just don't say anything more. Let them look at it and draw their own conclusions. So it's a very powerful tool to be able to show an image.
46:13Unidentified speaker (speaker_1): Yeah. I we're we're kind of growing the content team at Fiscal, and something I've told everyone is, you know, let let the graphics, let the pictures do the talking. Like you don't you you don't need to add copy to this. And it this one specifically actually ended up provoking a lot of engagement because, like you just said, the SP 500 I think was trading at right around all-time highs, but it felt like a lot of the constituents were not doing well. Like there was a lot of kind of carnage under the hood, but you had such good performance from big tech and
46:51Unidentified speaker (speaker_0): and
46:52Unidentified speaker (UNKNOWN): really
46:52Unidentified speaker (speaker_1): the top ten holdings that it was uh
46:57Unidentified speaker (UNKNOWN): it
46:57Unidentified speaker (speaker_1): was it was I I think timely was kind of why it provoked some such a response.
47:02Unidentified speaker (speaker_0): It's it's so true, and you know the content of this show is usually ever green and people are listening to it towards the end of 2026, six, probably around Thanksgiving, maybe the holidays. But uh I want to add that one of the feedback that I've been getting from listeners is that especially this year, on the surface, it's a bull market. Well, until recently it there's a bit of a pause. But they were telling me that it doesn't feel like a bull market to them. And a picture like this, I think says why, because there are quite a few stocks that are not hitting all time highs. They're actually at their lows and uh they they're very different from the handful of companies that people are following and paying attention to. Let's take a look at what we've been talking about earlier.
47:47Unidentified speaker (speaker_1): Insider buys oh it's between November sixth and November twelfth, we've got
47:54Unidentified speaker (UNKNOWN): I
47:54Unidentified speaker (speaker_1): don't have the company names here, so uh I'm looking at tickers, but we've got C CVS. V S. Uh looks like uh a executive vice president. president bought two million dollars worth of stock. KMI, fifteen million dollars worth of stock. That's pretty large. Um and it just it's it's prioritizing by value here. Then it says largest insider sales. Uh I think this is striker. Hundred wow, 179 million roughly. Um
48:27Unidentified speaker (UNKNOWN): and
48:27Unidentified speaker (speaker_1): yeah, uh I'm not gonna go through everyone, but you can see it's just basically prioritized by the value of the sale. And if you wanted, you could say, here, give me the uh give m give it to me in a PDF or whatever, so you can share it and email it, Slack it, whatever you want to do with it. Um or again, you could say, email it to me every Monday. Give me, you know, run it on a recurring time frame. Um But yeah. Yeah, it's it's a nice way to pull lists and and just extract data from kind of what would otherwise be pretty messy data sets.
49:02Unidentified speaker (speaker_0): It's very helpful to be able to see those transactions in a report and then have it delivered regularly if it's something that you want to pay attention to, just another data check in your process. I wanted to bring up one idea, kind of do a life little research. You know, we were talking about earnings. A lot of us are keeping up with our companies without any recommendation of any kind. Do you mind pulling up Google in your cloud and asking for the most recent quarter? What was it like? And I have in mind a few follow-up questions, but I just wanted the audience to see what. it's like to experience a conversation with the fiscal AI plugin when you get the data. And in my mind, once you see it, it's very hard to unsee it, as we said. It's very different than the experience of asking AI to source from all over, and you kind of have to do a sanity check if everything is accurate.
50:03correct, and really you know sourced properly. Here you get all fiscal AI data and you know where it came from. So you don't have to second guess what you're reading. It's uh coming from a reliable analyst.
50:18Unidentified speaker (speaker_2): Yeah, it's uh I thought
50:21Unidentified speaker (UNKNOWN): So what we're gonna do
50:22Unidentified speaker (speaker_0): we're gonna look at the last Google's earnings and then I have some follow-up questions about those earnings and maybe it will reveal a little bit more how I think about investment research at that level. But if you just ask Google the last earnings, just give me an idea, good, bad, worse, what's going on. And let's see how Fiscal with Claude's help can get. give us a look. So Ryan, you just showed us Google's earnings, just a quick look with some commentary. Tell us more what we're looking at and yeah.
51:08Unidentified speaker (speaker_2): Yeah, it starts with a verdict. So
51:11Unidentified speaker (UNKNOWN): it
51:11Unidentified speaker (speaker_2): says here revenue was a 127.6 hundred twenty seven point six billion. The beat uh that beat fiscal eyes estimate by just under one percent, driven by Google Cloud, growing 103%. Then it's gonna go each number results versus the expectation, uh variance, so how much did it beat it by? It's gonna go through the actual uh segment level numbers. We don't we don't yet have estimates. on segment level data, but eventually I think that's in the pipeline.
51:44Unidentified speaker (UNKNOWN): It's
51:44Unidentified speaker (speaker_2): gonna give the guidance. So said the fiscal year 2026 CapEx guidance was roughly 200 billion. Uh that was maintained. So it's giving the status essentially. How has the guidance changed? Cloud backlog that that guidance was raised. Foreign exchange, they expect to be worse. So it gets it gets pretty granular here.
52:11Unidentified speaker (speaker_0): What
52:11Unidentified speaker (speaker_2): drove the quarter? So just bullet points here in terms of what was best performers and then QA themes. This is good. This is exactly kind of what I was hoping they would touch on. TPU system economics So he asked what share of roughly $35 billion per gigawatt is booked uh
52:33Unidentified speaker (UNKNOWN): as
52:33Unidentified speaker (speaker_2): revenue and how much flows through to profit. Says we don't provide specific financials at the product level. Um he said this was the clearest dodge on the call. So it's kind of picking up, I think, tone for management and and whether or not they maybe avoid questions. But again, you can go directly to the conference call. It's going to give you a timestamp right there. Um TPU revenue said it was a relatively small portion of revenue this year, uh but a meaningful Q4 contributor with meaningful revenue ramp in 2027. twenty twenty seven. Anything stand out to you here? Bobmo?
53:16Unidentified speaker (speaker_0): So I I want to zoom in on one item. There was a non-operating charge that quarter. Can you ask about the non-operating charge?
53:26Unidentified speaker (UNKNOWN): So
53:26Unidentified speaker (speaker_0): it's usually one of those questions that would took a while to to look it up, see the details, see the commentary, what is it exactly? Has it been discussed in the transcript? And here you get a a quick rundown. of what it's all about and what kind of an impact it has. And from what I remember it has something to do with SpaceX.
53:49Unidentified speaker (speaker_2): Yeah, it says here the net was basically negative eight point two billion on the call. The CFO attributed it primarily to unrealized loss from our investment in SpaceX from a decline in its stock price in the third quarter. So I think coming off the IPO.
54:08Unidentified speaker (speaker_0): And just like a smart analyst, it goes a few steps ahead. So it gives you the exposure going forward, how much of SpaceX alphabet holds, and what kind of an impact it can have. So you can start asking questions and interact with something unusual, and then just like a smart analyst, it can kind of knows where you're going with it and adds more value. This is just one little example, but there are other companies that have something unusual or different in a particular quarter. And you know, we all try to remember everything, but I think it's super helpful to be able to look it up, ask one more question, go a little bit deeper, and maybe add a little brick to your you know conviction bank. or conviction beliefs here and allow you to hold it a bit longer if that's what you want to do, or you know, sell if that's enough for you to part with this company. But super helpful. Ryan, so many great examples. We went from m modeling
55:08to earnings analysis to seeing how you can build a whole report, how you can follow insiders and let letters and get a very full picture of an idea and go really, really deep. I love the graphics. I'm glad we had a moment to bring it up because with numbers, you know, we all look at things differently. I have spreadsheets that uh use colors, just like the one that you showed us. And uh they're, you know, I use conditional formatting, Forming, I love it. And just my brain thinks in colors. So when I look at a lot of data. I wanna see different shades of green or red, and I use other colors. That's how I process information. But I love that you can do it with fiscal AI and get a picture of the situation. And then you can as well remain silent and let the room do the thinking. But I I think it's very powerful.
55:59Unidentified speaker (speaker_2): Yeah, I think one thing that uh when I when I started using the MCP more and more, uh and I guess AI in general, the I was constantly thinking, can it really do this? Like like is it able to like whatever that task was, could it build a graphic that would be high enough quality the way I want it that I could post it? And just ask. Is is the advice I would give. Like, can you build a a three-statement financial model using fiscal IS data? And it Claude will kind of tell you what what would it think through? Am I capable of it? Yeah. Um, and so w whatever that workflow is that you're thinking about that's maybe kind of a time slog for you and and slowing you down, whether that's manual data entry, whether that's drafting reports for someone. Um
56:55Unidentified speaker (UNKNOWN): I
56:55Unidentified speaker (speaker_2): recommend testing out with AI plus an audited database. The other, I
57:01Unidentified speaker (UNKNOWN): guess,
57:02Unidentified speaker (speaker_2): tease I'll give here is the data that's available in the MCP continues to expand. So about a year ago, we launched the MCP and it was just fundamentals for US companies.
57:15Unidentified speaker (speaker_0): Now
57:15Unidentified speaker (speaker_2): it's fundamentals, KPIs, it's global coverage, basically now. Um, it's stock price return data, it's adjusted metrics, conference calls, fund letters, all the qualitative stuff. Soon, hopefully, I think in the next week, we'll also have macro data, which is a big launch and it's been something we've been working on behind the scenes for a while. So hopefully that can kind of provide extra context as well. Um, and it continues to expand and it just gets lumped into the MCP. There's no add on, no extra cost, anything like that. So um I recommend at a minimum giving it that free trial and and testing it out and seeing if it can improve your personal day day-to-day to day workflow.
57:59Unidentified speaker (speaker_0): We'll remind the audience that there's a discount available for listeners who It will be in the notes. You can look it up. And additional discount over the holidays, Black Friday, if you end up using it in a timely manner when this episode comes out. Explore it, see if it helps you. It's been a great blessing in my process, and it allowed me to improve my conviction on many fronts and sell an idea or two where I found something I wasn't aware of that evolved and changed. in a direction that I wasn't comfortable with. So it adds value in many different ways. It's just like having an analyst that you truly, truly trust. You know where the numbers come from. It's quick, it's efficient. You can have your own questions, you can have your own tools and see how it can enhance your process. And you know, the more people do fundamental research, I think the better the market can work because more. of us are actually looking at things, judging things, and deciding that this is worth holding. As much as passive investing has been a blessing for many people, we do need somebody that will question the truth and if it's good or bad, and if it's something that's adding value or not. We need people to do fundamental research. So the more tools they have, the more time they can spend actually thinking and generating. judging what comes out rather than, you know, coloring the
59:18graphics, but actually doing the work, I think the better for all of us. And as you said at the beginning, it's never been easier to do this kind of fundamental research. So if you love doing stock research, I mean you're gonna get lost playing with those things and including them in your workflow. Ryan, thank you so much. Any last thoughts for the benefit of this audience?
59:39Unidentified speaker (speaker_2): No, they thank you for having me. It this I I know it can feel intimidating. I like I was in the same boat and uh I I was kind of slow to adopt it in my workflows because I thought, you know, how much is it really, you know, I I don't want to outsource my critical thinking essentially to to AI. Is that gonna going to make me a worse investor? And I would say it's it's done the opposite. It allows me to spend more time on the critical thinking, more time around what's really going to matter for an investment and to to kind of allow the grunt work, but also have, like you said, sort of that customized, really, really smart analyst that's right alongside you while you're researching stocks. Anytime you have a question, just don't dump it in there and you can get answers, whether it's about a business, about its financials, about what other investors think of a certain development, it's it's really, really powerful. And you mentioned it, but on November 25th, I don't know when this episode's coming out, uh November 25th through the 30th, we will be running a 25% off sale on the MCP,
1:00:49Unidentified speaker (UNKNOWN): which
1:00:50Unidentified speaker (speaker_2): will be the first time we've done a discount. So if you go to Bogamil's link, you'll be able to find that 25% off sale.
1:00:58Unidentified speaker (speaker_0): Amazing. And again, MCP means fiscal AI plugin. Very easy. Added part of the same conversations that you're having with Shadow Cloud anyway. Just now you know where the data comes from. And it's just not just numbers, but as we said, quantitative and qualitative, everything that's out there that Fisco collected and serves to all of us, you know, great benefit. Thank you for being with me today, Ryan. What a blessing to have this tool. I wish I had this tool when I was 20 getting started. It would have saved me a lot of time. But anyways, I'm glad it's here now and helping me in my practice.
1:01:31Unidentified speaker (speaker_2): Yeah, thank you for having me. And I uh I hope your your audience got some value out of the discussion and can uh test it out for themselves.
1:01:40Unidentified speaker (speaker_0): You gotta see it. Thank you so much, Ryan. You were listening to Talking Billions. We talk about big ideas, big inspirations, big topics. We take on the hardest subject of all money. But our conversations lead us to an even bigger question: what it means to live a rich life beyond money. If you enjoyed the show, please take a moment and follow
1:02:02Unidentified speaker (UNKNOWN): the video.
1:02:03Unidentified speaker (speaker_0): subscribe, rate, and share with friends and family. We rely on word of mouth to promote the show. One click for you means the world to us. Thank you. Until next time, your host, Bogmel Baranoski.