Capital Horizons - The Podcast

Joe Wiggins argues that mean-variance optimization is flawed due to noisy inputs and advocates for robust portfolios focused on survivability and behavioral discipline.

52 min Listen to the original Episode page
Portfolio OptimizationBehavioral FinanceAsset AllocationRisk ManagementSystematic InvestingPrivate MarketsInvestment Process

Summary

Joe Wiggins explains that mean-variance optimization, once central to institutional asset allocation, is unsuitable as a primary decision-making tool because its inputs—expected returns, correlations, and volatility—are almost certainly wrong. He argues that optimization should instead serve as a sensitivity testing tool to understand how portfolios react to changing assumptions, rather than providing a single 'right' answer. Wiggins defines a robust portfolio not by maximizing ex-ante Sharpe ratios, but by maximizing the probability of good outcomes and minimizing catastrophic failure, emphasizing subjective judgment and scenario analysis. He highlights that an investment strategy's value is limited by the investor's ability to stick with it, suggesting 'survivability' should be a primary constraint in portfolio construction. Finally, Wiggins critiques short-term tactical allocation and advocates for systematizing processes to mitigate behavioral biases, recounting his own mistake of abandoning a systematic strategy during a drawdown only to see it succeed later.

Key takeaways

  1. 1
    Diversification provides two benefits: smoothing return paths and preventing behavioral mistakes during drawdowns by avoiding excessive concentration.

    Joe Wiggins · 13:31

  2. 2
    Systematizing investment processes removes in-the-moment judgment influenced by emotions or external pressures, serving as a primary defense against behavioral biases.

    Joe Wiggins · 33:15

  3. 3
    in understanding the long term. return drivers of an asset class is greater than it would be in predicting the short term return drivers of an asset class. In simple terms, I would say asset classes over the long run tend to be driven by fundamentals. Over the short run they tend to be driven by sentiment and technicals and and momentum.

    Joe Wiggins · 40:25

Episode sections

  1. 0:49 – 3:21

    Portfolio Optimization: Theory vs Practice

    Joe Wiggins explains that while mean-variance optimization was central to institutional asset allocation 10-15 years ago, its reliance on inaccurate forecasts and noisy correlations makes it unsuitable as a primary decision-making tool. He argues it should instead be used as a helper to understand how portfolios react to changing assumptions.

  2. 3:21 – 9:30

    The Illusion of Precision and Communication Challenges

    Speakers discuss how optimization creates an illusion of certainty that appeals to human discomfort with uncertainty. Wiggins notes that complex 'black box' models are difficult to explain to boards or investment committees, leading to the practice of constraining outputs to match pre-existing judgments.

  3. 9:30 – 12:54

    Defining and Building Robust Portfolios

    Wiggins defines a robust portfolio not by maximizing ex-ante Sharpe ratios, but by maximizing the probability of good outcomes and minimizing the probability of catastrophic failure. He emphasizes subjective judgment and scenario analysis over mathematical optimality, warning against quarterly rebalancing based on noisy inputs.

  4. 12:54 – 18:23

    Economic vs. Mathematical Diversification

    The discussion distinguishes between mathematical diversification (low correlation) and economic diversification (uncorrelated economic outcomes). Wiggins cites private equity/credit as potentially mathematically diversifying but economically correlated to equity risk, and notes that sovereign bond correlations change based on inflation regimes.

  5. 18:23 – 21:17

    Managing Underperformance and Behavioral Biases

    Wiggins argues that all models experience periods of underperformance, and investors often make rash decisions during these times due to outcome bias and myopia. He stresses the importance of setting clear expectations upfront regarding the duration and magnitude of potential drawdowns to avoid abandoning strategies prematurely.

  6. 21:17 – 31:48

    Survivability and Stickiness as Constraints

    The speakers agree that an investment strategy's value is limited by the investor's ability to stick with it. Wiggins suggests that 'survivability'—the ability to withstand underperformance without being fired or abandoning the strategy—should be a primary constraint in portfolio construction, potentially justifying lower expected returns for higher stickiness.

  7. 31:48 – 35:13

    Managing Behavioral Biases in Investment Processes

    Joe Wiggins argues that short-term horizons lead investors to focus on noise rather than signal, exacerbating behavioral biases. He suggests that institutional investors can mitigate these risks by establishing clear investment philosophies, systematizing processes to remove subjective judgment, and openly acknowledging group dynamics and individual biases.

  8. 35:13 – 40:06

    Systematic vs. Subjective Decision Making

    Wiggins recommends running a purely systematic version of an investment process alongside the subjective human approach. By comparing the outputs, investors can identify gaps caused by behavioral biases or valuable qualitative insights, though he notes humans often override models due to overconfidence.

  9. 40:06 – 45:50

    Time Horizons and Tactical Allocation Skepticism

    Wiggins distinguishes between long-term fundamental drivers and short-term sentiment/technical drivers. He expresses skepticism toward tactical asset allocation over short horizons, advocating instead for medium-term adjustments (3-5 years) based on valuation dislocations, while warning against predicting short-term market movements.

  10. 45:50 – 49:39

    Critique of Total Portfolio Approach (TPA)

    Wiggins views the Total Portfolio Approach as less transformative than advertised, essentially advocating for objective-driven allocation over traditional benchmarks like 60/40. He suggests TPA may serve consultant incentives for active/private market exposure and helps institutions avoid quarterly comparisons against passive benchmarks that they may underperform.

  11. 49:39 – 52:03

    Personal Investment Mistake: Abandoning a Systematic Strategy

    Wiggins recounts abandoning a systematic strategy combining trend following, equity market neutral, and risk parity after a period of poor performance. He sold due to daily monitoring and environmental pressure, only for the strategy to deliver stellar returns for the next five years, illustrating the cost of behavioral intolerance.

Claims worth verifying

Joe Wiggins is the Deputy Chief Investment Officer at St. James's Place Wealth Management.

Unidentified speaker (UNKNOWN) ·0:05

How to verify: Check the 'Capital Horizons' podcast episode description or St. James's Place Wealth Management executive team page. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

10 or 15 years ago, portfolio optimization was a central pillar in most institutional asset allocation processes.

Joe Wiggins ·0:51

How to verify: Review historical academic literature and industry white papers on mean-variance optimization adoption in institutional portfolios from the early 2000s to mid-2010s. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Mean-variance optimization uses expected returns, correlations, and volatility to find an optimal portfolio.

Unidentified speaker (UNKNOWN) ·1:17

How to verify: Consult standard finance textbooks (e.g., Markowitz, 2026) or CFA Institute curriculum on portfolio theory. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Return expectations and correlations used in optimization are almost certainly wrong over any forecasting horizon.

Unidentified speaker (UNKNOWN) ·2:02

How to verify: Analyze historical forecast accuracy studies (e.g., from the Journal of Finance) comparing forward-looking return estimates to realized returns. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Joe Wiggins previously worked at asset management firms where optimization models were used, and no one could explain why a 1% change in emerging market allocation occurred.

Unidentified speaker (UNKNOWN) ·4:41

How to verify: This is a personal anecdote; verification is limited to Joe Wiggins' professional biography and public statements from his previous employers. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Private equity and private credit may appear mathematically diversifying but are not economically diversifying to equity or public credit risk.

Unidentified speaker (UNKNOWN) ·14:16

How to verify: Review academic research on the correlation between private equity/credit returns and public equity/bond markets, particularly during stress periods. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

The correlation between equities and sovereign bonds changes based on whether the environment is inflationary or disinflationary.

Joe Wiggins ·15:00

How to verify: Calculate historical rolling correlations between equity indices (e.g., S&P 500) and sovereign bond indices (e.g., US Treasuries) across different inflation regimes. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Cliff Asness of AQR states that a model or strategy is only as good as your ability to stick with it through the long term.

Joe Wiggins ·23:08

How to verify: Search for AQR Investment Management publications or interviews with Cliff Asness discussing 'stickiness' or implementation shortfall. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Warren Buffett would have been fired multiple times during his career due to three-year periods of underperformance.

Unidentified speaker (UNKNOWN) ·30:33

How to verify: Review Warren Buffett's annual shareholder letters or biographies (e.g., 'The Snow Ball') for periods of significant underperformance relative to the S&P 500 over 3-year windows. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Investors significantly overestimate their ability to withstand drawdowns, confusing theoretical paper losses with lived experience.

Unidentified speaker (UNKNOWN) ·25:48

How to verify: Review behavioral finance studies on 'hypothetical vs. real' risk tolerance and investor redemption behavior during market downturns. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Joe Wiggins is Deputy CIO at St. James's Place Wealth Management.

Joe Wiggins ·31:48

How to verify: Check St. James's Place Wealth Management leadership page or LinkedIn profile for Joe Wiggins. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Primary source

Wiggins recommends a medium-term horizon of five years for valuation-based adjustments.

Joe Wiggins ·42:33

How to verify: Cross-reference with Wiggins' published articles or speeches on Capital Horizons podcast archives or firm publications. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Wiggins abandoned a systematic strategy combining trend following, equity market neutral, and risk parity after poor performance, which later performed well for five years.

Igor Yelnik ·49:39

How to verify: Verify via Wiggins' personal blog, LinkedIn posts, or interviews discussing behavioral lessons from past investment mistakes. For recent events, verify 2026-08-20 through 2026-11-20. Use another period only when the speaker explicitly names it.

Do your own research

  • Review academic literature on the instability of mean-variance optimization inputs (e.g., Ledoit and Wolf, 2003) to verify claims about forecast accuracy.
  • Examine specific portfolio construction frameworks that prioritize 'survivability' or 'drawdown control' over Sharpe maximization.
  • Analyze the performance of systematic trend-following or risk parity strategies during the period Wiggins describes abandoning his strategy to assess the cost of behavioral intolerance.
  • Evaluate whether 'Total Portfolio Approach' implementations at target institutions actually reduce benchmark-hugging or increase private market exposure as claimed.
  • Research the difference between mathematical diversification (low correlation) and economic diversification (uncorrelated outcomes) using historical data across inflation regimes.
  • Check for conflicts of interest if the speaker is affiliated with firms promoting private credit or equity as economically uncorrelated assets.
  • Review behavioral finance studies on 'outcome bias' and 'myopia' to understand the psychological mechanisms behind abandoning strategies during drawdowns.
  • Compare the performance of tactical asset allocation strategies against static allocations over 3-5 year horizons to test the skepticism regarding short-term prediction.
Full transcriptexpand

0:05Unidentified speaker (UNKNOWN): Hello

0:05Igor Yelnik: and welcome. It's Capital Horizons, the podcast where investment matters. I'm Igor Yelnik. My guest today is Joe Wiggings, Deputy Chief Investment Officer at St. James's Place Wealth Management. Hello, Joe, and uh welcome to the show.

0:21Joe Wiggins: Lovely to be

0:22Unidentified speaker (UNKNOWN): here.

0:23Igor Yelnik: We're going to discuss optimization and uh various uh behavioral biases which are prevalent in institutional investments today. Let's begin with uh how optimization works in practice. In institutional asset allocation process, uh how central in your opinion is portfolio optimization versus judgment and committee deliberation?

0:49Unidentified speaker (UNKNOWN): Yes,

0:49Joe Wiggins: good question. I think I that through my career the the importance of optimization has has varied. And certainly 10 or 15 years ago it was an absolutely central pillar in most asset allocation processes. And I've worked at a number

1:07Unidentified speaker (UNKNOWN): of

1:07Joe Wiggins: asset management firms where building long term SAAs was founded upon an optimization process. Basically lots of different ways of of approaching it, but in in simple terms,

1:17Unidentified speaker (UNKNOWN): most

1:17Joe Wiggins: are founded on some form of mean variance optimization where you're taking the expected returns for an asset,

1:23Unidentified speaker (UNKNOWN): um,

1:23Joe Wiggins: the correlation between those assets and the volatility of those assets, and then searching for an optimal portfolio given those inputs. And it's theoretically in that form very attractive because it's giving

1:36Unidentified speaker (UNKNOWN): you an

1:37Joe Wiggins: answer.

1:38Unidentified speaker (UNKNOWN): Uh

1:38Joe Wiggins: you've set put in a set set of assumptions, it's giving you an answer to what is the best portfolio, and the level of complexity in it also gives a lot of comfort to people as well, around the the robustness and inverted commas of of the investment process that led to the allocation. But it has a number of issues and issues that became very apparent to me as my career progressed. And

2:02Unidentified speaker (UNKNOWN): the

2:02Joe Wiggins: key to that is we're operating in the investment world, which has a a complex, noisy, adaptive system, which makes trying to optimize or trying to find the perfect answer to something very difficult because you have a set of variables that you put into an optimization which you know will be wrong. And certainly the return expectations or forecasts you put into an op optimization, they will almost certainly be wrong over the any horizon you're forecasting. The correlations will be noisy, the volatilities will probably be a bit more stable, but you're putting these inputs into a system, which you know will be wrong, and you know that the system you're using

2:41Unidentified speaker (UNKNOWN): will

2:42Joe Wiggins: be very sensitive to those inputs as well. So um the allure of that kind of s optimization process um and the theory of it probably outweigh outweighs the practice of the fact that optimization is probably the one framing for a noisy complex system. It's really we're trying to satisfy us or trying to find a sensible answer given the inputs we use are um likely to be unstable and at times inaccurate.

3:08Unidentified speaker (UNKNOWN): Um

3:08Joe Wiggins: but it's still still used and but I think probably hopefully being used in a different way rather than finding the one single portfolio answer, more being used as a tool to help um make a complex and difficult decision.

3:20Unidentified speaker (UNKNOWN): So,

3:21Igor Yelnik: in your opinion, this is just a helper, uh, this is not a primary decision-making tool.

3:28Unidentified speaker (UNKNOWN): Uh

3:28Joe Wiggins: for me, the best way to use it would be to aid your decision making if you're trying to work out what the the best allocation is over the long term, given the the objectives and constraints that you you might have. Optimizations are very useful, but useful in the sense that they can tell you what type of outcomes and paths that you might take in a particular scenario with a particular set of assumptions. But given none of us can predict the future, we don't have a crystal ball. We don't know what the returns for US equities or uh emerging market local currency bonds will be over the next 10 years. Um so we shouldn't optimize on the assumption that we know that. We should optimise on the assumption that we don't know that, which is really saying, okay, if these are my set of assumptions, what does that look like for a portfolio? And if I change my assumptions in a variety of different ways, how does that change the type of portfolio? And can I infer something from what different optimal portfolios look like when I shift and adjust these assumptions? For me, that's the best way of helping support judgment judgment-based based decision making. The idea that you can have a very complex optimization model that gives you the right answer

4:31Unidentified speaker (UNKNOWN): is

4:31Joe Wiggins: somewhat naive. And

4:33Unidentified speaker (UNKNOWN): I

4:33Joe Wiggins: remember one of my roles, there was an optimization model which was mean variance, but had some quite complex economic scenarios underneath

4:41Unidentified speaker (UNKNOWN): it.

4:41Joe Wiggins: And no one involved understood how the numbers were alighted upon. Um So it felt robust because there there's was lots of complex mathematical formulas underpinning it, but no one can do

4:50Unidentified speaker (UNKNOWN): it.

4:51Joe Wiggins: explain why your emerging market extra allocation had gone up by one percent compared to when you run it six months ago. Um so

4:59Unidentified speaker (UNKNOWN): there

4:59Joe Wiggins: are significant limitations when you can't explain something, particularly when um

5:04Unidentified speaker (UNKNOWN): as

5:04Joe Wiggins: I said, that we know that the numbers and our ability to accurately predict the the inputs is is limited.

5:11Unidentified speaker (UNKNOWN): I

5:11Igor Yelnik: think you have just raised uh two very important points. Uh the first one is that since we're dealing with the uncertain world, our forecasts are obviously wrong. They're always wrong. And when we optimize the portfolio, we get a precise answer. And the illusion is that this is the right thing because it's so precise, but it's just an illusion. And the second thing is that you create a complex algorithm. a complex model, which you still need to communicate somehow to the board or to the investment committee. And due to their responsibilities, they are supposed to understand uh what you are doing, what is actually hiding behind those complex models, what's the economic rationale? Eventually the decision is theirs. So how do you communicate those complex models and those complex solutions to the boards and committees?

6:05Unidentified speaker (UNKNOWN): Yeah,

6:05Joe Wiggins: that's a really great point. And the uncertainty one is critical. One of the reasons why optimizations in in theory are so attractive is because as humans we don't like uncertainty and we we struggle with uncertainty. We want to be confident and certain in any type of environment, particularly in financial markets. And

6:24Unidentified speaker (UNKNOWN): as

6:24Joe Wiggins: a somewhat as a facade, optimizations say actually. this is the right answer. This is the efficient or objectively best portfolio and given the optimization process we followed, which we know is a a fanciful concept to follow, but it does make us feel a bit more comfortable about an uncertain world and a a future that's inherently unpredictable. So behaviourally I can definitely understand why the level of complexity and the level of inserta level of certainty that's implied in optimization to so attractive. On the communication point, it it's incredibly important, both from a when discussing the boards or committees about why you're making ass allocation changes, you know, it's rather clear investment thesis that you can assess through time. And it's impossible to assess the quality of your judgments or decision-making decision making through to time an if you can't really explain why you made that choice in the first place. And certainly in my career, there's been situations where optimization processes have been used and the optimization has

7:22Unidentified speaker (UNKNOWN): come

7:22Joe Wiggins: up with answers that were not palatable for whatever reason or led to odd allocations across different portfolios. And then what happens is

7:29Unidentified speaker (UNKNOWN): we

7:29Joe Wiggins: don't really like these answers. So we're going to apply some judgment and put some constraints on the optimization so it comes up with the answer that we're more comfortable with. So you still have that this

7:38Unidentified speaker (UNKNOWN): is

7:38Joe Wiggins: an optimal. optimized portfolio, but then you just put constraints around it to come up with an answer that you're more comfortable with. So you have this um this idea or concept that it's a very optimized portfolio, but you're just using that optimization to get you to the the constrained portfolio that you're that you're comfortable with. Um I think the un underlying that type of behaviour behavior is the fact that black box complex optimizations where you can't clear. clearly explain the rationale for allocations or why allocations might change just lead to a range of really difficult decision-making problems through

8:10Unidentified speaker (UNKNOWN): time.

8:12Igor Yelnik: Okay, but uh when you when you have this uh optimized portfolio and then you uh tweak the weights to make it feel comfortable, don't you discredit the whole idea of optimization?

8:27Unidentified speaker (UNKNOWN): Um

8:27Joe Wiggins: I think you acknowledge the limitations in optimization when applied to the real world, which is what which is why I think that any optimization process which blindly takes inputs and then accepts the outputs just doesn't work and certainly isn't something I would consider using now. So it optimization in that in that form is

8:45Unidentified speaker (UNKNOWN): is

8:45Joe Wiggins: limited. That's why back to your earlier question using optimization optimizations is not about finding the one right answer, it's about assisting judgment-led decision making and answering questions that are part of the investment process, such as if I adjust my return assumptions by X, what happens to the optimal portfolio state, or if I assume stock bond correlation over the period I'm interested in in changes compared to what was the case in history, how much does that change? the allocation to bonds in my portfolio? So answering those types of questions is useful and helpful. I think it's less helpful in answering the ultimate question about what is the exactly right allocation you should have in your portfolio given your objectives.

9:30Igor Yelnik: Yeah. But eventually you want to arrive at uh at some kind of robust portfolio. What is robust is for everyone. to interpret. And I have seen the different interpretations of the idea of a robust portfolio. How do you define a robust portfolio?

9:49Unidentified speaker (UNKNOWN): So

9:49Joe Wiggins: I tend to think about it as the portfolio which you believe maximizes the probability of delivering good outcomes and minimizes the the probability of of delivering very bad outcomes. And you can tailor that depending on specifically what your objectives are and how you might frame very bad outcomes. So I don't think robustness should be viewed in terms of maximizing an X ante sharp ratio or necessarily maximising a return. But thinking about plausible future scenarios for the world and for asset class return and behaviour and trying to rely upon the allocations that you think are are robust across those different allocations. and avoid scenarios where um you fall very very short of your return objectives. But within all of this,

10:35Unidentified speaker (UNKNOWN): I don't

10:35Joe Wiggins: think there's a there's a there's a right mathematical answer that gets you there. There's a a significant amount of subjective judgment that leads you to to make the call and what you think a a sensible allocation is. And again, from a behavioural behavioral perspective, because the system is so noisy and adaptive, uh an appropriate approach where you're you're satisfying and think what you're doing is good enough, or based on the information and knowledge you have, is probably better than trying to find an optimal solution to that problem.

11:02Igor Yelnik: Right. So practically uh there is some kind of trade-off between optimality and let's call it stability

11:10Joe Wiggins: of

11:10Igor Yelnik: the portfolio. What kind of trade-offs uh would you be willing to explain? accept between optimality and stability? And what is unacceptable?

11:20Unidentified speaker (UNKNOWN): I

11:20Joe Wiggins: think for me it's it's about the magnitude of the change involved and your confidence in the in the inputs that has led to that change. Um so if you are if it's a if it's a minor change in, for example, in the return assumptions and it's leading to very significant changes in the optimal portfolio because the optimization process is very sensitive to those inputs. Then you have to question whether the magnitude of that change is sensible relative to the confidence you can possibly have in those in those forecasts. Um,

11:52Unidentified speaker (UNKNOWN): and

11:52Joe Wiggins: you also just need to understand how significant is the is the level of change

12:00Unidentified speaker (UNKNOWN): in

12:00Joe Wiggins: the portfolios. Um

12:02Unidentified speaker (UNKNOWN): so

12:03Joe Wiggins: making quarterly changes based

12:04Unidentified speaker (UNKNOWN): on

12:05Joe Wiggins: new optimal portfolios um is probably just going to be portfolio noise. And if there are significant changes made based on um material alterations in the assumptions you're using the build portfolio, then that becomes um more viable. But but you have to have thresholds around the confidence in your in your inputs and the and the magnitude of the changes. If the magnitude of the changes doesn't look meaningful, then it's unlikely that the that moving to a new preferred portfolio allocation is likely to be worthwhile.

12:35Igor Yelnik: How about diversification? It's quite clear that in order to build a stable portfolio, uh you need to diversify. The the idea of diversification is built on independence or at least uh decorrelation of returns. But one thing is to model it and another thing is you know to put it in production and see your portfolio develop in real life. How do you tell uh the real diversification from the diversification which also only exists in the model?

13:04Unidentified speaker (UNKNOWN): Yeah,

13:05Joe Wiggins: it's a good question. I I think diversification for me comes in two forms. One is the that diversification being used in

13:11Unidentified speaker (UNKNOWN): in

13:11Joe Wiggins: terms of the the the types of outcomes you can get from a portfolio of over your over your investment horizon. So one asset class outperforming another means there's a benefit to owning some of that asset class and your uncertainty around which out asset classes will perform the best over that period means it makes sense to be diversified on on the basis we can't predict the future. And then the path of return matters, so being diversified appropriately can smooth the path of returns and particularly avoid behavioural mistakes. through time if you are sensitive to periods of of drawdown and then diversification can allow you to get into a position where you're not making poor decisions because you're too concentrated in a particular asset class that's performing poorly at any any given point in time. What I think the most important thing to do is to to complement any assessment of let's call it mathematical diversification with the economic rationale for diversification. So, why should what should the relationship be between two asset classes? And how should we think about that relationship in terms of how we model it or how we build a portfolio? So, a couple of examples of that might

14:16Unidentified speaker (UNKNOWN): be

14:17Joe Wiggins: private equity or private credit. Mathematically may look diversifying, but economically is not genuinely diversifying to equity market risk. or public credit risk that you have in a portfolio. It just appears to be diversified. So you might get some of that smoothing that I talked about, but you're not genuinely diversifying your economic outcomes or only doing it at the at the margin. So that can be diversification that is overstated mathematically compared to economically. And another example is um and probably relevant to lots of market conversation today is about the diversification properties of nominal sovereign bonds or nominal treasuries, and lots of talk around

14:59Unidentified speaker (UNKNOWN): the

15:00Joe Wiggins: relationship between equities and sovereign bonds and how that correlation relationship has has changed and bonds not being as an effective diversifier. To me, I would argue the bonds have just been behaving in exactly the way you'd expect them to behave in an inflation a more inflationary. regime. And it's not that bonds have changed, it's just that the environment or backdrop has changed, so the correlation relationship between bonds and equities is altered. And that needs to be something you think about in your building your portfolio. So the the relationship between equities and bonds will be different in a disinflationary environment to one where there's more inflationary pressure. And when you model out different economic scenarios and apply process. probabilities to that that should be affect that should be reflected in how you might size your bond allocation. I think really difficult to predict which economic regime we might be in. And it's more about thinking there's probabilities attached to different regimes that might affect my assumptions about relationships and and correlations. But getting diversification right and doing in a sensible way is is crucial critical from a behavioural perspective. and from an investment outcomes perspective.

16:05Igor Yelnik: Which uh probably also means that a longer time series uh of returns is not necessarily a good indication of what you should expect in the future. And uh, even though technically you have uh higher statistical confidence in your outcomes, in practice you are just drawing from a distribution which is no longer in existence. Like if we try to apply the correlation structure of let's say 1990 to

16:31Unidentified speaker (UNKNOWN): uh

16:32Igor Yelnik: 2020 uh to today's environment, it's probably not going to work perfectly well. The most obvious example is the equity bond correlation, which uh which has changed sign.

16:45Unidentified speaker (UNKNOWN): Yeah,

16:45Joe Wiggins: I think that's right. So taking one path of an asset class relationship from the past and assuming that that is the stable or average level for that relationship going forward is a very dangerous thing to do with with certain asset class relationships. So that's why I think it's important to try and understand what is the economic rationale about why two asset classes should have a certain type of relationship or should perform in a certain way relative to each other, and why should they be diversifying to each other. So again, when you look at private equity versus public equity, why should two asset classes that are in essence investing in the same underlying things be diversifying to each other economically? They they shouldn't be. It's just because you price them differently that they they behave differently. So understanding why are they diversifying, how confident you might be in that diversification and and for some people that diversification is useful, but the smoothing of returns and the lower volatility it has a utility attached to it but often it's marketed in a way that it's more diver genuinely diversifying than it actually is. So actually trying to sit back and say

17:52Unidentified speaker (UNKNOWN): what's

17:52Joe Wiggins: the underlying rationale for the relationship between these asset classes, how might I expect them to behave in different environments is is incredibly important.

18:01Unidentified speaker (UNKNOWN): Yeah

18:02Igor Yelnik: indeed. However, whichever model you choose, you choose one model. Yes, you can tweak its forecasts after the fact, but nevertheless, it's some kind of model, and you work out of that model. And then inevitably when you put it in production, there will be periods of underperformance. And then the big question is whether the model is wrong, or is it simply going through a period of underperformance? How do you think about it? How do you tell the difference?

18:31Unidentified speaker (UNKNOWN): Yeah.

18:32Joe Wiggins: Good question. All models are wrong. Some some are useful. I I think it's it's important to set expectations correctly. So w whatever investment approach we take, whatever model we might use to inform our investment decision making, we all go through significant periods of looking like we are wrong or incorrect about our view even if we make very good decisions we'll go through periods when that decision looks wrong and that and that creates challenges because um a you'll doubt yourself and begin to question every aspect of of the model and difficult to tell in that moment whether it's a a broken model or just a model behaving in a in a normal way through

19:13Unidentified speaker (UNKNOWN): the

19:13Joe Wiggins: cycles of performance that we never have with any asset class or any model or any any fund manager for that for that matter. So that is that is a challenge. We tend to make really bad decisions under performance pressure as well. So when performance is poor we tend to make rash short-term outcome-led decisions um even if we have the best intentions of making very sensible process led decisions. But Well I think that for me the key around

19:38Unidentified speaker (UNKNOWN): dealing

19:38Joe Wiggins: with with performance and periods of underperformance for any for any aspect of investment is to set expectations correctly at the outset.

19:47Unidentified speaker (UNKNOWN): So

19:47Joe Wiggins: people react very badly to to surprises when things behave in a way that's very different to how they might expect. And this is why investors when they performance chasing as they tend to do in mutual funds get very surprised when they buy a fund when it's performed incredibly well for three years and it goes on to underperform and they're they're shocked that it can go through periods of underperformance and then inevitably end up selling it at the worst possible time. So you need to set expectations very correctly for for any model and say, given the structure of this model, in the event that we are we are correct in terms of the what the model provides us and how it works, what what is it reasonable to expect in terms of periods of underperformance, both in terms of the duration of that underperformance and the magnitude of that underperformance. Because I think people often think about drawdowns of underperformance from a magnitude perspective. Whereas

20:37Unidentified speaker (UNKNOWN): it's

20:37Joe Wiggins: almost harder when it's over a long period of time, when those doubts grow over a long period of time about whether what you're doing is is working anymore. Or But being clear about those expectations at the outset and comparing the reality with those expectations is important, behaviorfully, but also in terms of how you might assess the robustness of any given model.

20:58Igor Yelnik: Yeah, uh you're of course right about investors uh selling uh their investments at the worst possible time. However, most optimization frameworks assume that investors will always faithfully implement and maintain their allocations recommended by the models. In this slide. uh should uh we treat investor behavior itself as some kind of constraint in portfolio construction?

21:23Joe Wiggins: Yes. So when we are able to place sensible constraints around our behaviour, particularly when if we can make judgments about the future when we're in a cold state and avoid making decisions about things when we're in a hot state in the moment and maybe emotional or making rash judgments in the future, that is a very good way to guard or protect against some of our worst behavioural biases. I think putting parameters within models that account for the fact

22:01Unidentified speaker (UNKNOWN): that

22:02Joe Wiggins: we will make certain types of decisions um

22:05Unidentified speaker (UNKNOWN): that

22:05Joe Wiggins: in the moment feel right, but over time are likely to be suboptimal is a very good way to approach it. So knowing ourselves as investors, knowing how we're likely to behave in certain conditions and trying to create models that um place sensible limits or restrictions on on some of our worst worse behaviours, behaviors I think is a sensible thing to do. And that clearly much or only possible to do before it happens, in the event then we definitely won't put any constraints on ourselves when we're feeling under under pressure or when we're um beholden to a particular bias.

22:37Igor Yelnik: Yeah. And uh looking at it from another side,

22:41Unidentified speaker (UNKNOWN): uh

22:42Igor Yelnik: like imagine that we have used our more sophisticated model with all the constraints, all the uh regularization techniques and shrinkage techniques that we could possibly apply. But uh the portfolio which we believe is optimal as a result of this model's work is not really optimal uh if the investor can't stick with it, right?

23:05Joe Wiggins: Yeah, I think it's um

23:07Unidentified speaker (UNKNOWN): Think

23:08Joe Wiggins: Cliff Asnes at That AQR who talks about the

23:11Unidentified speaker (UNKNOWN): a

23:11Joe Wiggins: model or investment strategy being only as good as your ability to stick with it through the long term. So it needs to be shaped by the by the personal behavioural reality of how you might act, and also the institutional reality if you're working with an organization. Um how willing is that organization to sit through or withstand certain levels of underperformance or certain time periods of underperformance because it's all well and good you you're choosing a fantastic model that will deliver over the long term, but you get fired after two years because it's underperformed. Um it's probably not a great model. So s survival matters a lot. So how you build survival into the shape and framing of a model is incredibly important. I always say that the most important thing for an investor to think about whether it's select an asset class fund or whether designing a model is survival, and can you survive the potential return? paths that might come out of a particular investment choice?

24:05Igor Yelnik: Yeah, which is a question which not too many investors ask themselves, although they should. Another angle to the same question basically: if we have two portfolios which have equal expected returns, equal risks, but one is much easier for investors to stick with through a difficult period, and another one is more difficult. Probably these portfolios should not be regarded as economically equivalent.

24:34Unidentified speaker (UNKNOWN): I don't

24:34Joe Wiggins: know that's true. No, I would say

24:36Unidentified speaker (UNKNOWN): that

24:37Joe Wiggins: most investors, even if the returns weren't commensurate, if the returns were lower for the portfolio that was easier to hold through time, there's probably still a case for that being the preferred portfolio because the survivability point becomes incredibly important and we're not comparing theoretical ex ante returns here. We're comparing what returns

25:01Unidentified speaker (UNKNOWN): is

25:01Joe Wiggins: an investor likely to receive over time investing in this portfolio accounting for behaviour. There's So some which might which probably feels quite ephemeral, but there is some behavioural cost or behaviour behavioural reality that you need to apply to the path of a portfolio's returns. And higher theoretical returns are great, but not if no one ever ever receives them. So I think to an extent of you'd accept a a lower forecast return from a portfolio that you're much more confident you can hold on to through time than and a higher return from one where the probability of you making poor choices or abandoning that portfolio is notably higher.

25:42Igor Yelnik: Is there any way to quantify this stickiness through difficult periods?

25:48Unidentified speaker (UNKNOWN): I

25:48Joe Wiggins: wish I could. It's incre incredibly difficult. I think that um everyone hugely overestimates their ability to withstand those periods. I think what a lot a lot of people do is that they'll consider in essence a a paper drawdown. They might stress test or run scenarios on a portfolio and they'll say, well, in this type of environment I might lose ten percent or I might lose fifteen percent or underperform the market by a seven percent. And looking at numbers on

26:17Unidentified speaker (UNKNOWN): a

26:17Joe Wiggins: piece of paper on a computer screen in theory is just entirely different. That's not how human behaviour behavior works. Human behaviour behavior is about the lived experience of going through those spells, going through those spells either as an individual investor on your own, going through it for an environment where you're at work, you're checking your portfolio every day, you're going to report to committees on a monthly or quarterly basis. You're getting lots of criticism about the performance. But I think, although it's really difficult to to quantify, I think organizations or institutions will probably have historical precedent for. how they've reacted to periods of underperformance through time. So looking at situations where they might have had funds or portfolios or models that haven't been working, they can probably have a sample of situations where they've abandoned funds or portfolios or investment strategies. And they can infer from that what their tolerance might be for periods of of challenge or or difficulty. Um

27:15Unidentified speaker (UNKNOWN): but

27:16Joe Wiggins: hard to hard to apply a number to it. But any investment team or firm or or business should be asking themselves,

27:23Unidentified speaker (UNKNOWN): how

27:24Joe Wiggins: much tolerance do we think we've got for

27:26Unidentified speaker (UNKNOWN): under

27:26Joe Wiggins: performance? And that might they might have investors may have significantly more tolerance than their committee has or their clients have as well. So, in essence, your tolerance is as much as the the the weakest weakest limit. link in the stakeholder chain that you might have. Um, doesn't matter if you've got tolerance for it, if your clients don't, doesn't matter if you've got tolerance for it if your investment committee doesn't. So you've got to understand all different stakeholders and what their tolerance is. And um, and if you want to increase the tolerance of people involved and stakeholders involved, then you need to communicate with them and set expectations correctly about the

28:01Unidentified speaker (UNKNOWN): the

28:01Joe Wiggins: realities of the investment. strategy you might create.

28:05Unidentified speaker (UNKNOWN): Yeah,

28:05Igor Yelnik: it's it's a very good idea to try to measure your own tolerance, at least based on your past track record of of intolerance. Probably it could work. I'm not sure if anyone has done that.

28:16Unidentified speaker (UNKNOWN): Yeah,

28:16Joe Wiggins: and I think the when when people make those decisions, um when we when we uh abandon strategies that aren't working, it it won't be because

28:26Unidentified speaker (UNKNOWN): the

28:26Joe Wiggins: it's just our tolerance to under performance. It'll be because something is broken in the process or the model's wrong. Um so hard for us to look back and say, yeah, really we abandoned this 'cause performance was poor and we couldn't we couldn't take the pain or pressure, pressure from that. But the really good thing for firms to do is try to try and audit the decisions they've made and just look at well, okay, what's the average level of underformance before we've capitulated on an the idea or strategy.

28:51Igor Yelnik: Yeah. Speaking of various behavioral biases, when you read academic literature, there are like hundreds of them uh listed there. Probably they can be reduced to a smaller number and uh to an even smaller number when we think about the biases which matter most in the real real-world world portfolio management. in real-world portfolio decisions on an institutional level. So which ones do you think are most important?

29:23Unidentified speaker (UNKNOWN): I

29:24Joe Wiggins: think to the to the optimization problem and question that we've we've talked about already, overconfidence is a is a significant one. So overconfidence in the accuracy of the forecasts that we make. So our willingness to make forecasts. about uh the unpredictable future and imponderable variables is is higher, and we generally think that we're better than we are, and think we're better than other people, and ignore the fact that other people can't seem to make these forecasts, but but we can, and that has significant implications for the investment decisions we make and the portfolio allocation choices that we make. We are very significantly biased towards

30:02Unidentified speaker (UNKNOWN): our.

30:02Joe Wiggins: outcomes rather than process. So that outcome bias means that we see noisy outcomes in in the investment world and infer that something is either working or not working or is good or bad. And as we already said, we know

30:16Unidentified speaker (UNKNOWN): that

30:17Joe Wiggins: any investment strategy through time will go through good periods and bad periods. That the problem is is 'cause because we're so biased towards outcomes that when things are going wrong, we'll say that the process is broken and we need to to change it rather than it being normal fluctuations in in value. We always say

30:33Unidentified speaker (UNKNOWN): that you look

30:34Joe Wiggins: at Warren Buffett's track record, he's probably the um the greatest investor of his generation uh on the time horizons of most investors and the behavioural patterns of most investors. He'd have been fired on multiple occasions through his career for those three year periods of of underpower. underperformance because during those three years we'd be saying that he's lost it as an investor and

30:56Unidentified speaker (UNKNOWN): it

30:56Joe Wiggins: no longer works as a as a strategy. So that bias towards outcomes um being incredibly important is

31:04Unidentified speaker (UNKNOWN): um

31:04Joe Wiggins: endemic and incredibly powerful. The other one I'd say is

31:08Unidentified speaker (UNKNOWN): is

31:08Joe Wiggins: myopia, so being incredibly sensitive to short-term results at most short-term results. in and short-term outcomes in financial markets are just noise, random noise and variability in in returns. But what we tend to do is overweight them and we'll tend to attack some narratives around them as well and explain why these things have happened when ultimately, as I say, it's just random fluctuations that shouldn't be a a significant feature in how you make assessments of your in investment approach. But like we see on a day-to-day and a week-to-week basis, just an incessant desire to assess every fluctuation in markets over short horizons, which not only is completely exhausting, but it just means you're focusing on noise over signal for the vast majority of the time. And so then the more you can dial that down as in. an investor and focus on what matters, um, the better. But an incredibly difficult challenge and and probably one that's getting harder and harder in the world we're in now, where we are just bombarded with with noise about investments and people can trade very easily and very freely and often with leverage, um, in quite a quite a straightforward way.

32:21So I think um

32:23Unidentified speaker (UNKNOWN): the

32:23Joe Wiggins: ability of investors to make poor decisions and to react badly to their behavioural biases is is greater than it's ever

32:31Unidentified speaker (UNKNOWN): been.

32:32Igor Yelnik: Is there any way on an institutional level to insulate the portfolio, maybe make by make some changes to the investment process from those behavioral biases?

32:44Joe Wiggins: I think

32:46Unidentified speaker (UNKNOWN): the

32:46Joe Wiggins: the key to it is, and this there's this is definitely not a a foul fail self-face. self faith this is definitely not a fail safe method, but the the key starting point is to have

32:55Unidentified speaker (UNKNOWN): a

32:55Joe Wiggins: clear set of investment principles or philosophies, something that underpins what you think about markets and that should incorporate behavioural considerations as well. So some assessment of what you believe about investor behaviour, what you believe about your own behaviour. behavior. If you don't have a clear philosophy, then it's impossible to have a process because why are you building a process? if you don't believe anything about financial markets? And then having a clear process, clear, well defined, hopefully largely objective process. Um and elements of that process that you can systemize back to the earlier part of our conversation, the more you can systematize an investment process and take it away from in the moment judgment based on how. you might be feeling or the pressures that might be applied to you at any given time. I think that's that's very important. Communication is also really important and being open and honest about behavioural biases, calling them out in advance, acknowledging the fact that we're going to experience it them

33:55doesn't stop you making really poor choices because of them. But if you're open about we're all susceptible to them both as a group as individuals, that can help you

34:06Unidentified speaker (UNKNOWN): at

34:06Joe Wiggins: least mitigate some of the most damaging aspects of of investor behaviour. And they're important to think about the fact

34:13Unidentified speaker (UNKNOWN): that

34:14Joe Wiggins: we often think about behavioural bias as being an individual phenomenon and often an in individual phenomenon experienced by other people, not us. But the other point to note is group dynamics have their own set of behaviour behavioural issues as

34:28Unidentified speaker (UNKNOWN): well, um,

34:29Joe Wiggins: which are similar in some ways and different in other ways. Um so you need to think about not just how I will make decisions, but this group of people that are involved in decisions, how do they interact as a group, what are the biases that they might hold as a group and as a collective, and how might that differ to um the operations of individual biases?

34:50Unidentified speaker (UNKNOWN): Okay.

34:51Igor Yelnik: But if we think about different possibilities to insulate the process from wrong decisions, can we think about a quantitative model which might potentially protect in the investment community from uh those behavioral biases? Or do you think that people simply will find ways to override the model uh when it is in conflict with their beliefs?

35:13Unidentified speaker (UNKNOWN): They

35:13Joe Wiggins: will uh they will definitely do that, but but it's still a useful process. But one thing that I've always liked to do, and this is part of how we think about our own asset location, is to say

35:26Unidentified speaker (UNKNOWN): if

35:26Joe Wiggins: we were to

35:27Unidentified speaker (UNKNOWN): build

35:28Joe Wiggins: our asset location process but run it in an entirely systematic way, so no no judgment apart from the point you create in the model, purely systematic, we want to run that. alongside our pure approach, which includes our own subjective judgments and is a noisier, inherently noisier process.

35:48Unidentified speaker (UNKNOWN): And

35:48Joe Wiggins: then just run that systematic process and that will come up with its own recommendations, choices, allocation decisions based on our philosophy and process, but divorced from any of the noise that we might be personally experiencing. And that is a useful way of saying actually, we're making this decision, but this systematic process that we've made, which tries to replicate our process

36:09Unidentified speaker (UNKNOWN): in

36:09Joe Wiggins: a quantitative fashion, is actually doing something in a very different way. Why is that? So it's a useful way for asking questions and trying to find what is the gap between the model we might build, a systematic version of what we might build, and then the judgments that we're actually making. And part sometimes it might be actually there's some nuance and some colour that we get from our own qualitative judgment that adds to decision making. Other times it might be behavioural biases that has drawn us away from what we actually think our process and philosophy should be. So I think that's a useful way of doing it, as you suggest that there is no perfect way of creating a structure that stops us making um behaviourally behaviorally poor or jeopardizing. jeopardized decisions. But I think comparing it to one

36:55Unidentified speaker (UNKNOWN): where

36:55Joe Wiggins: there's very limited noise and where it systematizes a useful way of of protecting yourself.

37:00Igor Yelnik: From what I have seen, uh it very often happens that you have a model, you put it in production, and uh then there are two possibilities. It is The first one is that people don't like what the model suggests. And they tweak the forecasts of the model, sometimes just to feel comfortable, sometimes simply because they don't like it. And another thing is that when they accept the models uh forecast, the model suggestions, and things don't work as well as expected, uh, then uh people can use the model as as a a scapegoat.

37:38Joe Wiggins: Yeah I think that's right. And and even when when we when we review and look at systematic investment strategy so systematic equity um mandate for example

37:51Unidentified speaker (UNKNOWN): a

37:51Joe Wiggins: lot of the questions will be around okay when do you override that model in what situations

37:56Unidentified speaker (UNKNOWN): um

37:56Joe Wiggins: will you use your judgment to overlay the model if it's periods of market stress when we think that the the inputs into the model might be uh at risk of being inaccurate or skewed. Um but people are generally more worried about b the model being left to its own devices than they are the human judgment becoming involved in and overriding the model which is which is an interesting interesting dynamic. Um But people

38:25Unidentified speaker (UNKNOWN): I

38:25Joe Wiggins: think you'd you're generally more skeptical about the the pure output of a of a model than they are than their own judgment, even if the underlying inputs and the philosophy and process supporting the decisions are are consistent.

38:42Unidentified speaker (UNKNOWN): Yeah,

38:42Igor Yelnik: that's probably because of uh human overconfidence, which you mentioned before.

38:46Joe Wiggins: Yeah, absolutely.

38:47Igor Yelnik: Well well we' we w we like to think that's that we're way better than the average, and of

38:50Unidentified speaker (UNKNOWN): course

38:50Igor Yelnik: better. than the stupid machine.

38:53Unidentified speaker (UNKNOWN): Yeah,

38:53Joe Wiggins: absolutely. And and when interesting when I I've spent a lot of time speaking to to fund managers and often discretionary fund managers who are doing kind of stock picking style approaches and and when you ask them about okay, if you were gonna replicate your approach but do it systematically, how would you do it? And that they often just balk at the entire idea of running it systematically and saying it's impossible 'cause impossible because they've got this secret source.

39:14Unidentified speaker (UNKNOWN): To

39:15Joe Wiggins: which I would say, okay, fine, but then just just do it and see see what comes out and compare it to your judgment and see what the what the gap is and then you can more clearly articulate what the difference is between the the pure systematic model and the human judgment and that the

39:29Unidentified speaker (UNKNOWN): um

39:29Joe Wiggins: the gap between those two things could be very instructive, whether that's a gap that is beneficial because you're adding new information that the model can't pick up or might be negative because you're allowing more behavioral biases to affect the decision making.

39:43Igor Yelnik: Yeah, indeed. Uh now optimization methods are used for both strategic and tactical allocations. And human biases also work very differently on the strategic and tactical time horizons. So what can be done to prevent uh technical tactical views from destabilizing the strategic portfolio?

40:06Unidentified speaker (UNKNOWN): Yeah,

40:06Joe Wiggins: so I think it's it's very important when you take asset allocation views away from your longer term view that there are there are parameters that are set around it to to limit the level of confidence you might express in um

40:20Unidentified speaker (UNKNOWN): in

40:20Joe Wiggins: that view. I'd I would say generally speaking, my my confidence

40:24Unidentified speaker (UNKNOWN): and

40:25Joe Wiggins: in understanding the long term. return drivers of an asset class is greater than it would be in predicting the short term return drivers of an asset class. In simple terms, I would say asset classes over the long run tend to be driven by fundamentals. Over the short run they tend to be driven by sentiment and technicals and and momentum. I personally am very sceptical about tactical asset ass allocation location approaches generally. uh insofar as they they tend to be about people trying to predict how markets will move over short time horizons, which I don't think anyone can do,

41:00Unidentified speaker (UNKNOWN): um,

41:01Joe Wiggins: because the system is just too complex, too noisy to make regular, consistent decisions that add value. Um so the way w I would think about it is

41:12Unidentified speaker (UNKNOWN): we

41:13Joe Wiggins: have a might have a long long-term term portfolio. portfolio allocation driven by our SAA over the medium term. There might be dislocations in markets over, let's say, three to five years. There might be significant valuation dislocations, which mean we want to make adjustments to those. But those adjustments need to be conditioned or sized based on your confidence in getting those calls right and the consequences of being, well, we're making those making those judgments. um the shorter our horizon is um I think the

41:46Unidentified speaker (UNKNOWN): the

41:46Joe Wiggins: the the greater the risk of us being the wrong sides of very unpredictable market market movements. So I would be um generally very wary of any tax class asset allocation location approach that suggests it can can predict the market. And

42:01Unidentified speaker (UNKNOWN): I think

42:01Joe Wiggins: you may want to complain complement your very long run asset location approach with something that is a medium term horizon where you

42:09Unidentified speaker (UNKNOWN): can

42:09Joe Wiggins: lean into opportunities on over the over the medium term that might arise from a valuation perspective. Um but again you do not want to do so in a way that significantly jeopardizes how you built the portfolio to deliver over the long

42:24Unidentified speaker (UNKNOWN): run.

42:26Igor Yelnik: Interesting.

42:27Unidentified speaker (UNKNOWN): Uh what

42:28Igor Yelnik: is medium term for you when we s whether you speak about technical asset location?

42:33Joe Wiggins: So I would say five years is medium term for me. My my general view on markets is that there are two two things that matter

42:42Unidentified speaker (UNKNOWN): and

42:42Joe Wiggins: two different approaches you can take. You can take a valuation led approach,

42:47Unidentified speaker (UNKNOWN): um,

42:47Joe Wiggins: where you really care about the price of an asset and the fundamentals of the asset, and that's that tends to work over longer horizons so five years plus um they say three to f three years plus maybe uh or you can take a momentum based approach and you you're more interested in sentiment and and price trends which you want to adopt over say say a one year period. Obviously there's strategies which will try and combine both of those things together but I think most things in markets can be put down to to valuation and and momentum and they work over over different time horizons.

43:17Igor Yelnik: Yeah. That's interesting because as you know, the strategy which we run at our firm is uh basically a global tactical estilocation strategy. We call a systematic macro, but in essence it's a global tactical estilocation strategy. It's quite slow because it's easier to make uh early return champ forecasts uh for longer longer periods of time. Uh but uh it's not 30 years, of course.

43:41Unidentified speaker (UNKNOWN): Yeah,

43:42Joe Wiggins: absolutely. I think the one of the important things is your horizon needs to match your investment approach. So

43:49Unidentified speaker (UNKNOWN): the

43:49Joe Wiggins: the the best investment approach you might adopt needs to be tailored towards your investment horizon. If you ask someone to make choices over the next year, the variables that you might consider will be very different to if you ask someone to make a judgment about the next. ten years. Um so your your time

44:05Unidentified speaker (UNKNOWN): horizon has

44:06Joe Wiggins: to have the next five minutes indeed. So your your horizon has a huge impact on the variables you care about and the strategy you wish to adopt. And you get you get lots of noise in markets with people debating preferences and philosophies when all that is is they've got very different horizons, so naturally they're going gonna to have different approaches to markets.

44:25Unidentified speaker (UNKNOWN): Okay.

44:26Igor Yelnik: Now all portfolio optimization frameworks are based on some assumptions.

44:30Unidentified speaker (UNKNOWN): Uh

44:31Igor Yelnik: if there was uh just one assumption that you could change, which is commonly embedded in most frameworks, what would that be?

44:40Joe Wiggins: I I think that the the the main challenge I've had with with optimizations is the the the precision of the of the return forecasts. Um I think too often

44:53Unidentified speaker (UNKNOWN): the

44:53Joe Wiggins: return forecasts are

44:55Unidentified speaker (UNKNOWN): very

44:55Joe Wiggins: precise when we know with the vessel in the world that our forecasts will be wrong and they're probably wrong by a significant magnitude and often even our rankings of asset class returns will be wrong over over time as well. So having a an optimization process which is very sensitive to those return forecasts and those return forecasts being very specific rather than being

45:17Unidentified speaker (UNKNOWN): a

45:17Joe Wiggins: range, maybe with probabilities attached to them, I think creates a significant flaw in traditional optimization approaches. I mean, we've seen lots of studies around SAA and strategies, return forecasts that have been made, and how inaccurate those forecasts are, even over the over the long run. So the more you create optimization. processes which are sensitive to being right about those return forecasts, and the more that brings into question the validity of

45:47Unidentified speaker (UNKNOWN): that

45:47Joe Wiggins: optimization process.

45:50Igor Yelnik: What do you think about the uh total portfolio approach, which is now being promoted by some Um pension funds?

46:00Joe Wiggins: I would say that it's not as different or transformative as as some people are making out. Um in essence, really what I think it's saying is it's important to have the right um benchmark or the right objective, which is tailored towards the the goals of whatever plan you're you're running, and then to design your portfolio around those objectives by assessing the characteristics of the available asset class or opportunity set. I mean, that seems quite similar to other ways of allocating portfolios or building portfolios. Um, I think what it does tend to do is move you away from let's call it the traditional 60 40 benchmark, um, and probably brings a bit more flexibility in how you allocate assets compared to those more traditional structures, which might be overweight or underweight, uh, an extra XT bond split. for example. Um

46:53Unidentified speaker (UNKNOWN): but

46:54Joe Wiggins: also aware

46:54Unidentified speaker (UNKNOWN): that it

46:56Joe Wiggins: does serve a couple of incentive purposes for consultants, for example, who generally speaking

47:02Unidentified speaker (UNKNOWN): would

47:02Joe Wiggins: like to advocate for asset allocate for more exposure to active strategies and more exposure to private markets. Um the TPA TBA approach, which you're more divorced from that market benchmark, allows you to do that.

47:13Unidentified speaker (UNKNOWN): And a

47:14Joe Wiggins: lot of strategies. a lot of asset location approaches have underperformed that simple passive 60 40 style approach for a number of years. So removing that from the regular uh quarterly update that you give to your investment committee could be an attractive option as

47:29Unidentified speaker (UNKNOWN): well. I

47:30Joe Wiggins: honestly I think it's

47:32Unidentified speaker (UNKNOWN): there

47:32Joe Wiggins: is a spectrum of approaches um from a like very rigid SAA based approach. approaches where you don't deviate much from a 6040, and then much freer approaches where you might have a real return objective and you build a portfolio based on the the attractiveness of any given asset class and their relationship to each other. Um and I think different institutions will sit somewhere on in different places on that on that spectrum. But

47:59Unidentified speaker (UNKNOWN): to

48:00Joe Wiggins: me it doesn't feel like anything particularly new, and you can could easily run what looks like a very SAA driven approach through um a TPA lens as well when really what you're doing is assessing the yeah the attractiveness of the of the opportunity set and trying to build a sensible portfolio on that basis.

48:18Igor Yelnik: Don't you think that the biggest difference between uh SAA and TPA is who makes what decisions? Because uh in in TPA, um more decisions are delegated to the investment team, and uh the board is not responsible or the investment committee is not responsible for the strategic stalkation anymore. They just set uh set a benchmark.

48:44Joe Wiggins: Possibly, although I think it's

48:47Unidentified speaker (UNKNOWN): and I've

48:47Joe Wiggins: said I've read this a few times, but I think it's it's a bit of a mischaracterization to think. that the SAA approaches are naturally set by

48:53Unidentified speaker (UNKNOWN): a

48:54Joe Wiggins: committee that can be an investment team who thinks the best way to s to build a portfolio and to start building a portfolio is with an SAA with an SAA approach. Um so I I I think

49:05Unidentified speaker (UNKNOWN): you

49:06Joe Wiggins: could easily create a TPA approach that had committee oversight and you could run a process that was more traditionally SAA oriented that didn't have a committee involved in it. at all. So it feels like that part is um

49:19Unidentified speaker (UNKNOWN): the

49:19Joe Wiggins: governance oversight and the decision making

49:23Unidentified speaker (UNKNOWN): is

49:23Joe Wiggins: definitely interesting and and definitely important but I don't n necessarily think it should define the approach that's adopted.

49:31Unidentified speaker (UNKNOWN): Okay.

49:32Igor Yelnik: And my final question to you is about your personal experience as an investor. what has been your biggest mistake that you have ever done?

49:46Joe Wiggins: Um my biggest mistake, this significant list to go through. Um

49:53Unidentified speaker (UNKNOWN): I

49:53Joe Wiggins: recommended uh uh some years ago, I recommended a systematic strategy which combined um a number of different styles of systematics. systematic. So there's some trend following in there, some equity market neutral, and some risk parity was being used in the the portfolio as a as a as a diversifier. Um started off performing very well and then went through a really horrendous performance performance spell when lots of systematic strategies were were struggling in trend following was in one of its barren spells. Um ended up abandoning, capitulating on the strategy, in part because of the environment I was in, in part because I was checking performance every day. And it was pretty painful getting punched in the face looking at performance every day. And it got too much to bear and then inevitably ended up selling it. Everyone was delighted that it was no longer in the portfolio. And then it shot the lights out for the next five years with stellar performance. Um which was a very good behavioural lesson if from for a variety of angles about going back to her earlier point about behavioural tolerance and the the environment you're in and assuming that things are broken because they came through periods of of not working. So um not a

51:06pleasant experience but one that has hope hopefully shaped Shape me reasonably well as an investor moving forward from

51:12Unidentified speaker (UNKNOWN): there.

51:13Igor Yelnik: Yeah that's a good story.

51:16Unidentified speaker (UNKNOWN): Uh

51:17Igor Yelnik: Joe, thank you very much for this interview. Thank you very much also for the interview that you contributed into our book on uh perspectives on portfolio optimization.

51:27Unidentified speaker (UNKNOWN): Uh

51:27Igor Yelnik: we're expecting this book to be

51:30Unidentified speaker (UNKNOWN): uh

51:30Igor Yelnik: published in the first quarter of next year. It's currently in production. Our first book in this perspectives perspective on series has seen the light of the day already, and as far as I understand, it's been some success. So I hope that the second one will also be quite successful. So thank you very much and goodbye.

51:52Unidentified speaker (UNKNOWN): Thank

51:52Joe Wiggins: you very much. Good luck with the

51:53Unidentified speaker (UNKNOWN): book.

This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.
All insights