Announcer: Welcome to TDAM Talks, a TD Asset Management podcast. Join us for insights and analysis on current themes in capital markets from our thought leaders. From market insights to investment strategies, we'll help you navigate the complex landscape of investing.
Ingrid: By definition, quantitative investing is a sophisticated approach to investment management, security selection, and portfolio construction. It uses a combination of computer models, data and statistical analysis, and professional judgment to achieve an outcome, be it passive return, lower volatility, or in today's conversation, excess returns or alpha. It also focuses on creating those optimal portfolios that balance risk with return. My name is Ingrid Macintosh, and on the podcast today, we're going to be talking about quantitative investing, and with a quantitative team that drive the strategies at TD Asset Management. Joining me today is Julien Palardy, head of Quantitative and Passive Investing here at TD Asset Management, and Philip Gendreau, Vice President on the quantitative team, collectively managing over $15 billion in quantitative equity strategies. Gentlemen, welcome.
Julien: Thank you, Ingrid.
Philip: Yeah, thanks for having us.
Ingrid: OK, so I'm hoping that this podcast is one we can keep coming back to. So I really want to get a little academic right out of the gates. How did I do in my definition of quant investing? And maybe help our listeners understand what and what the benefits are of using quantitative investing?
Julien: I think that your definition was pretty good. I would say that, over the last maybe 20-- maybe even more than 20 years-- 30 years, there's been quite significant increase in the amount of science that has been applied to finance, and more specifically, investment management. It probably started with Markowitz really, and then it spread out outside of just portfolio construction and model construction.
And now I would say that quantitative investing is really about maximizing the use of science and scientific approaches to the art of managing money. And that doesn't necessarily mean that being a human being is out of the equation when it comes to quantitative investing. But the role of the human being is quite different. Obviously, there's the role of building those models that ultimately are used in applying those quantitative strategies. But it's also, let's say, more of a risk control type of role as opposed to stock beginning. Not sure what'd you say, Phil, if you share the same view.
Philip: Yeah, I think you really nailed the definition. I would just add-- well, as you said, we're really trying to make investing as systematic as possible, really trying to remove emotion from the process. And hope ultimately, over time that [AUDIO OUT] outcome.
Ingrid: It's interesting, because I think when-- we've been working together for many, many years. And we often think about fundamental versus quantitative as art versus science. But really, a lot of the same fundamental vital factors are incorporated in the investment process, but in a quantitative way. So can we talk a little bit more about that distinction between fundamental and quantitative, where they're different but where there's similarities as well? Because I think that'd be really helpful for our listeners
Julien: I would say, yeah, I agree that, at the end of the day, we looked at roughly the same numbers. The way we looked at them is going to be different. In our case, it tends to be very data-driven. So we know those numbers every morning, and we crunch them. Probably a fundamental performance measure will have specific companies that they will tend to follow, and they are going to have a really deep dive into those [? stoppings. ?]
But they also tend to have a general view about what the future looks like. While in the case of quant strategies, these are machines. At the end of the day, they lack imagination. To be honest, that's probably the weak spot of quant strategies. However, they're really good at crunching a lot of numbers across the entire investment universe. This is where we get our edge really.
Philip: Yeah. And I don't think it's either or. You don't necessarily have to choose between the fundamental way of doing things or the quant way of doing things. We target the same outcome in the end, which is [INAUDIBLE]. But just because we look at different things, we have different ways of looking at things, we're probably going to outperform or underperform at different times. So by, I think, both ways of doing things, you can also gain the benefit of the diversification.
Ingrid: Yeah, a really complementary approach. And I love the way that you talked about that piece about the screening or where a fundamental portfolio manager starts versus the ocean that the quantitative manager can start with, and really start screening down to those outcomes. We're going to get to the big A question in a moment. But let's start more fundamentally with a conversation around technology. How are we using technology to make better decisions? And then how does that really work within the traditional portfolio manager-led decision-making?
Philip: As we just said, in quant investing, most of the decisions are driven by models. And we rely on a lot of data for our models, so technology is very important to what we do. We process daily, actually, millions of data points, anything from the company's fundamental data to training cost estimates. And other than working with the data, there's a few other different ways we use technology in what we do. So in quant investing, we rely heavily on portfolio optimization. Basically, that's just a fancy way of saying we build portfolios with the help of the computer.
And with portfolio optimization, we can really control the risk in our portfolios, either total risk, like for our mobile strategies, or active risk in the case of the alpha-driven strategies. We can target specific factor exposures, such as dividend yields in our dividend portfolios. We can also keep trading costs low by re-trading where there's the most liquidity, and so on.
Ingrid: It really is a matter of determining the outcome you're trying to get to. Because while we're going to be focused on the alpha strategies today, TD Asset Management was a Canadian pioneer in low volatility investing, where the specific outcome we were looking for was benchmark, like returns, but with less volatility in support of our pension fund clients and their risk budgeting process. And what I'm hearing you say is, again, depending on the factor you're trying to control for, the outcome you're trying to engineer, you can adapt the models to that.
You led off by talking about very, very large amounts of data. So of course the next question I'm going to ask you is about AI. What does AI do for quant investing? How is it changing the landscape? Can you talk a little bit about that? Because I think everybody's talking about productivity and what AI can and can't do. But I think quant investing is a great place to start with that conversation.
Philip: So the main way we can use AI in quant investing is it allows us to process a lot more information. So we're able to explore new data sets that we really couldn't use before. So just to give you an example, a few years ago, when we were using financial statements from companies, we really could only look at the numeric data-- so [INAUDIBLE] statement, balance sheet, and so on. Now, with the help of new AI techniques, like natural language processing, we're able to go beyond the numbers and use the actual text in the financial statements, as an example, or the earning call transcripts, really try to understand what the margins are saying beyond the numbers...
Julien: And I'm not sure if you remember. But Phil is referring to financial statement data. But when we launched those strategies, that wasn't there. That was more than 45 years ago. We were getting this data on floppy disks at this stage. It was probably taking a month before it could get loaded and crunched through our models. And now it's instantaneous. Those numbers get released, and we add them in our model. So it's not only the amount of data that has increased massively, it's also the speed at which we get this data and at which we can process this data. And ultimately, it gives an-- that's the edge of quantitative investing. Phil was talking about the amount of data. Just as a reminder, computers nowadays, they beat human beings-- most human beings at chess. And chess is only about 64 data points that you need in real-time to make your decisions. And technically you can build a winning strategy just with those 64 data points. For finance and the markets, every morning, our models, they process millions of data points across investment [INAUDIBLE]. And Phil was talking about optimization. To get the perfect portfolio, you literally need to look at all those data points and the covariance matrices and figure out what portfolio released today and the least volatile. You can't do this on the back of a napkin, as you can imagine. So if machines are pretty good at chess, they're even better at making complex decisions in that case.
Ingrid: I mentioned at the outset, statistical data analysis, loads of data, and also some human judgment. If it was just data, and everybody had access to the same data, one would believe, when applied in the same way with the same rules, people would be replicating each other's quantitative strategies. So can you talk a little bit about the art of creating a quantitative strategy? How is it that you land on the factors that you use to inform the model? Because people are thinking, well, if it's not human, it's not human. Can anybody do this? Maybe talk about that factor design and how you come down to strategies.
Julien: Actually, I'm going to start by addressing the first thing you said about the [INAUDIBLE] of quant strategies. You'd be surprised, given the same amount of data, how different strategies can be. And in fact, we see it at market. The same event happens and two or four managers will have a vastly different interpretation of what's going on. So same thing when you build quant models, you're going to have the same data set. And there's going to be different models that will be built around the same data sets, and different investment decisions that will emerge from this. And as you multiply the amount of data, you can imagine that you also multiply the number of models sets and the number possible decisions that can be generated by those models. And Phil Is building models day in and out, so he can tell you about how he approaches this and things like factor selection and how he works with factors. Phil can maybe--
Ingrid: Maybe talk a little bit about backtesting too, like how you land on those factors.
Philip: Even if everything we do is very systematic, it's not 100% science. We're humans building models, so we have to understand that we're going to introduce certain biases in the models. And it's really important to understand what we're betting on in the end. Because it's not, because it's systematic that really there's no biases. You just asked about that backtesting. Especially for backtesting, it's important to make sure, for example, that there is no forward-looking bias because it's very easy to produce a good backtest. Live performance is something else. So what we're trying to do is always to have the backtests that are as representative of the future. As a team, we can make that. Because it doesn't do us any good to have a good backtest. We want good live performance.
Julien: This is, by the way, what makes the difference between beginners and people who are actually experts at this and who can actually build models that work live. So we have plenty of people that pretend [AUDIO OUT] amazing backtest,
Ingrid: But they don't apply.
Julien: But those models, it's very different. Yeah.
Ingrid: I want to talk a little bit about market environment. And while I'd say that I want this podcast to be evergreen and sort of principles-based in discussing how we do what we do, we're recording this mid-late 2024, at a time where there are a number of variables causing a much higher degree of volatility in the markets, or unknown variables that are affecting the markets. Can you talk about quant investing in this environment, where it helps it, how it manages for these non-standard shocks to the system, if you will, and how it's playing out?
Julien: Yeah. By the way, very good timing for this podcast because I don't think you could have picked a better time to discuss this. So over the last maybe year and half, we're seeing an increasing amount of concentration in the markets. [INAUDIBLE] see that for most human beings, we're humans, so driven a lot by emotions, and obviously, guiding the markets when you see a handful of stocks generating most in the performance.
Ingrid: Yeah, like the Mag Seven phenom we've witnessed, right? Yeah.
Julien: Exactly. It's easy to live through for a couple of weeks. But after a year and a half, I mean, a lot of people just make the decision to buy those [AUDIO OUT] that name-- sorry, but-- and minimize their underperformance. And in fact, if you start a portfolio, if I asked you to build S&P 500 tomorrow with a handful of stocks, probably you're going to start with those names, right. And the rest, you're going to build around that. And you know what? This leads to overweighting those names. And quite often, in the recent period, it would have led to our performance. So it's a natural tendency for human beings to do this. When you run a quant strategy, you do the opposite. You have a view on every single second in your investment universe.
And by default, those Magnificent Sevens are just going to be a handful of those names on which you have views. You may have the right views on them, or maybe not. But at the end of the day, it's very unlikely that you're going to end up overweighting each one of those names. So you're going to have a natural tendency to underweight them, which can lead to underperformance. As long as you stick with the process, eventually things will reverse. And this is exactly what we've seen in the last few weeks. Yes, if you're a human being, it's easy to go in the other direction and say, OK, I'm going to cut my losses and build a strong position or go benchmark weight or even overweight on those names. If you follow the process, and if you're disciplined, then you're going to stick with the process.
And this is what generated reversal in performance, at least a partial reversal in performance, that we've seen in the last few weeks. So this applies to both the disciplined alpha strategies, the low volatility strategies as well. And unfortunately, what we see in the market is that people have a time limit where you're willing to outperform. After that, they decide to change their approach. And sometimes it's in the worst possible time they choose to do this.
Ingrid: Philip, you made the comment at the outset where you talked about strategies being complementary, quantitative and fundamental, because they behave a little bit differently. In a longer time stream, can one or either of you talk a little bit about, generally speaking, in which environments do quant strategies do better?
Obviously, it's a long-term proposition, investing. But in the shorter-term, what are the types of market environments that are really good for quant strategies? And which are the ones that are kind of challenging and you would expect them to not be performing maybe where we would want them to?
Julien: Yeah. I'm going to come back to what I just said a second ago. Essentially, one big thing that can move against the quant managers is when, again, you have a view on, let's say, the S&P 500. It's going to be 500 individual names. Or it can have [AUDIO OUT] average you're going to be right, possibly, hopefully.
But even if, on average, we're right, in terms of what means will outperform and underperform, in the recent market conditions, you would have to be right on a handful of names. If you're wrong on those, probably you're going to miss the target, and you're going to underperform. So this is a type of environment where it's clearly more difficult for quant managers, while fundamental managers may have a very, very deep view into those handful of names. And they can figure out which ones are going to be, let's say, Teslas, and which ones are going to be Nvidias.
But to achieve this, you need to have, not only a deep understanding on the name, but also you need to understand what's beyond the numbers, right? It cannot just be about the fundamentals. It has to be about the future of self-driving cars. It has to be about the future of GPUs as well, and the future demand, and where the prices are going to go and thinks like this. It requires some degree of imagination that machines may not have entirely yet, I would say.
Ingrid: Yeah. As I hear you, I think, in periods where highly concentrated markets are really doing well, quant strategies may lag. But in those moments of reversion, the quant strategies are probably going to be faster to that recovery or to that reversal. I want to shift our conversation a little bit because part of the rationale for having this podcast at this point is we are on the verge of a renaming, if you will, or a rebranding of some of our active strategies. So we've just recently renamed the TD US Quantitative Equity Fund, the TD US Disciplined Equity Alpha Fund, and will be launching a TD Global Disciplined Equity Alpha Fund in the fall. So we've created this family under the banner of disciplined equity alpha. Can you talk a little bit about the reasons for the name change? What's common about the strategies? What's different?
Julien: OK well, the name change is very easy. It's marketing. So ideally, when you have a product, you want the name to reflect what this product is about. Simple view. If you have a cake, you want to call it delicious cake, because it's supposed to be delicious, right? You don't want to name the cake how it's made, right? You don't want to build in the recipe in the name of the cake. So originally, for the US Disciplined Alpha [INAUDIBLE], the original name, I know a lot of people know, it was TD US Quantitative Equity. So quantitative, that's how we get there. It doesn't tell when it's going to deliver. So guess what. No [INAUDIBLE] is also quantitative. Right?
Ingrid: Yeah. Different outcome by design, yeah.
Julien: Exactly. So disciplined alpha is exactly what you're going to get. It doesn't really matter if it's quantitative or fundamental. It's going to be disciplined alpha, which means we're going to try to deliver you alpha, and it will be consistent as possible over time. So excess returns-- that's the objective. And now, when it comes to US versus global, we have the expert here. So, Phil, maybe you can say a few words about the differences and commonalities as well.
Philip: Yeah, sure. Well, the US is about 60% to 70% of the global investment universe, so we already have a pretty long, like 15-plus years track record in the US market. So really, the new global funds, that's going to be like a 100% new. Let's call it 2/3 something that we're already used to doing. And then the other part, for those who want exposure to other markets, they're going to get in this fund.
And so in the US, we really have a model that looks at the entire market. But in the global space, because we believe that there are different return drivers by region, a separate model for each of the regions in our investment universe. So usually we would divide the global landscape in five regions. So we believe, with this approach, we can add models that are more targeted to the specificities of each regions.
Julien: And as Phil says, so the US is 2/3 already of the investment universe, and we're that for a long time. And guess what. That's the most difficult 2/3 as well, because the US RF GAAP is by far the most efficient market in the world, in my opinion, and I'm pretty sure in Phil's opinion. So I'm not going to say that the rest is a piece of cake, but--
Philip: It's easy.
Julien: Let's say it's a piece of delicious cake.
Ingrid: [LAUGHS] And it is an important distinction, right? Because you're absolutely right. Our investors, our advisors that we partner with, they do have a huge landscape to select from. They've got a variety of fundamental managers where they're looking at portfolio managers. And then, once you go to quantitative, it can mean passive. It can mean low-vol. In this case, we're talking about alpha. Can you talk a little bit about the objectives for this strategy? How much return are you trying to generate for a given level of risk? And maybe talk a little bit about how you bake that delicious cake, if there's anything that you can share in the recipe.
Julien: Yeah. So I'm going to tell you how delicious it's supposed to be. While the target is-- we're always a bit reluctant to talk about this. But before fees, we should expect something in the range of 1.5% because we have a 4% active risk budget, as we call it. But we rarely take that, so it's typically lower. So our return objective should be roughly in that range, which is similar to what we target in the US.
But more importantly, what we want-- and that's the delicious part-- but we want to deliver that ideally in all types of market conditions. We know that there's going to be circumstances where it's going to be more difficult, but we want this to be as consistent as possible over time. That's why we qualify ourselves quite often as being style agnostic. So we don't want to do well only when [? the world ?] does well, when value does well. We ideally want to figure out what's going to work well, and then we position ourselves consequently. And now the how-- the cook is here, so.
Ingrid: Without giving away the recipe, but general ingredients.
Philip: Yeah, well, so basically, the portfolio is built around the benchmark. So we don't want to go too far in either direction. It'll probably be value or growth. We're trying to stay around the benchmark and take-- [INAUDIBLE] that's where we believe we can add value. So that's where we were talking before about the portfolio optimization. That's where it really comes into play. To build a portfolio, that's basically only hitting from the benchmark where we believe we can add value.
Ingrid: So then, when advisors or investors are thinking about making an allocation to quantitative investing, what's it doing for their portfolio? What would you say to investors about a disciplined alpha strategy in their portfolio?
Julien: It really depends on what the advisor is looking for. And to be honest, there are so different views out there as to what they're looking for. So it depends what their clients are looking for as well. So obviously, this one would be for people who have clients that look closely at what the markets are doing, and they want to be able to deliver better returns as consistently as possible over time.
So obviously, the assumption here is that, outside of equities, you're going to have bonds. You're going to have possibly alternative asset classes also to compensate for the risk of equities or to mitigate the risk of equities. So risk withing the equity space has to be a secondary consideration, as opposed to a primary consideration. And return should be your primary consideration for a fund like this. And it's really meant to be, at the core, someone's equity portfolio. Because as I was saying, this as a strategy then is style agnostic. So in some circumstances, it may be growth biased. And in some circumstances, it could be value biased. So you cannot even use in history to figure out where it's going to be going forward, because the models are going to adapt. So it's really meant to be a [? quart ?] refund nearly as a replacement for downside exposure or an ETF exposure.
Ingrid: And, Philip, you mentioned the risk budget. So you're really trying to deliver that incremental return with no surprises, no significant downside shocks to an investor's portfolio.
Philip: Well, relative to the benchmark.
Ingrid: Yeah, yeah. I understand that, absolutely. That's the one thing-- of all the things we can control, that's the one that we can't control. Thank you so much, gentlemen. Anything else, any final thoughts, as we're looking at probably the back half of '24 right now, in a pretty volatile environment ahead? Any final thoughts?
Julien: I'd say stay invested. And stay safe, as well.
Ingrid: That's my line, guys. [LAUGHS]
Julien: Oh, sorry, sorry.
Ingrid: That's how I close. That's awesome. OK, gentlemen. Thank you so much. And to our listeners, I hope what you've taken away from today is that quantitative investing isn't just one thing. It can be a variety of strategies designed to deliver specific outcomes. And specifically today, we've been talking about how quantitative investing can deliver that incremental added value to your portfolios in a consistent and predictable way. So, gentlemen, thank you so much. Again to our listeners, for more information like this, you can follow us on Spotify, Apple, and Amazon. You can find us at tdassetmanagement.com or follow us on LinkedIn. Thanks, everybody. And as Julian said, stay safe.
Julien: Thanks, Ingrid.
Ingrid: Thanks, guys.
Announcer: The information contained herein is for information purposes only. The information has been drawn from sources believed to be reliable. The information does not provide financial, legal, tax, or investment advice. Particular investment tax or trading strategies should be evaluated relative to each individual's objectives and risk tolerance. This material is not an offer to any person in any jurisdiction where unlawful or unauthorized. These materials have not been reviewed by and are not registered with any securities or other regulatory authority in jurisdictions where we operate. Any general discussion or opinions contained within these materials regarding securities or market conditions represents our view or the view of the source cited. Unless otherwise indicated, such view is of the date noted and is subject to change. Information about the portfolio holdings, asset allocation, or diversification is historical and is subject to change. This document may contain Forward Looking Statements, or FLS. FLS reflect current expectations and projections about future events and/or outcomes based on data currently available. Such expectations and projections may be incorrect in the future, as events which were not anticipated or considered in their formulation may occur and lead to results that differ materially from those expressed or implied. FLS are not guarantees of future performance, and reliance on an FLS should be avoided. TD Global Investment Solutions represents TD Asset Management Inc. and Epoch Investment Partners Inc. Those entities are affiliates and wholly owned subsidiaries of the Toronto-Dominion Bank.
[MUSIC PLAYING]
DISCLAIMER:
The information contained herein is for information purposes only. The information has been drawn from sources believed to be reliable. Information does not provide financial legal tax or investment advice. Particular investment tax or trading strategies should be evaluated relative Each individual's objectives and risk over this material is not an offer to any person in any jurisdiction where unlawful or unauthorized.
These materials have not been reviewed by and are not registered with any securities or other regulatory authority in jurisdictions where we operate. Any general discussion or opinions contained within these materials regarding securities or market conditions represents our view or the view of the source cited. Unless otherwise indicated. Such view as of the date noted and is subject to change.
Information about the portfolio holdings asset allocation or diversification is historical and is subject to change. This document may contain forward looking statements or affiliates for less reflect our expectations and projections about future events and or outcomes based on data currently available. Such expectations and projections may be incorrect in the future, as events which were not anticipate or considered in their formulation may occur and lead to results to differ materially from those expressed or implied.
FLS’ are not guarantees of future performance in reliance on FLS’ should be avoided. TD Global Investment Solutions represents Asset Management Inc and Epoch Investment Partners Inc., both entities are affiliates and wholly owned subsidiaries of the Toronto-Dominion Bank.