Insights


Podcast: Inside the Mind of a Quant, Part 2

In this podcast, Olivia Engel, CIO of Active Quantitative Equity, speaks in depth on how quantitative investing seeks to exploit market inefficiencies by extending human judgement, experience and expertise in a uniquely powerful way, using data science.


Olivia Engel
CIO, Active Quantitative Equity

Podcast Image

Podcast Transcript

Celina. Welcome to “Human Led, Research Tested,” the Active Quantitative Equity Podcast. I’m Celina Rogers. This episode is part two of my conversation with Olivia Engel, Chief Investment Officer for Active Quantitative Equity here at State Street Global Advisors.  Last time we discussed the philosophy that underpins quantitative investing, including the behavioral biases that equity market participants are subject to, the influence these biases can have on stock prices in the short term and the way these forces interact with longer term fundamentals. Today we go into greater depth on how quantitative investing seeks to exploit market inefficiencies, by extending human judgement, experience and expertise in a uniquely powerful way using data science.  We picked up the conversation talking about how different market players have different investment goals and objectives.

Olivia. Every market participant has a different objective. Some have a very short horizon in which they’re viewing the holding period for a company. Some have a very, very long horizon and some may be just seeking to take advantage of a single earnings announcement with a belief that it will be better than what the market was expecting. Others are taking advantage of a disruptive secular shift in an industry’s whole existence and seeking to take a long term position based on that belief. And so, because you have all these different market participants acting with different objectives in mind – that's what makes, I think, the market so interesting and able to be exploited in different ways.

Celina. So there’s a formalized construct in which you identify these long-term and short-term factors and understand how they interact with each other.

Olivia. Yes, exactly. So some of the signals that we will use to forecast stock returns are centered around longer-term phenomena, and they tend to be valuation- and quality-based indicators. And then some of them are shorter term, and they tend to be revolving around investor behavior and investor sentiment towards companies. And because trends tend to persist for some period of time but don't go on forever generally – whereas valuation can take some time for the market to realize. But ultimately over the long term the market generally does realize true value. The length of time it takes to realize value is variable across different industries, different market environments, but on average you know it can take sort of twelve months to two years for true value to be – for perceived value to be realized. And then indicators that are more sentiment based generally can have a payoff of around three to six months. So investing on a trend, for example, can have a very good payoff for a three- to six-month period.

Celina. There is a sort of formalized framework of investment themes in AQE. Value is one, quality another, both longer term, and sentiment.

Olivia. So value, quality and sentiment are the main themes that we think are very important from a bottom up, core characteristics that we’re seeking in companies. Now the way we do research has actually other dimensions to it, because it’s not just about picking the companies that will have the highest returns. It’s also about how you put them together in a portfolio, how you trade in and out of different companies, what your buy and sell discipline is, and then comparing your forecast of returns with what it might cost you to trade. So transaction costs also need to be thought about and researched. But most of our research is oriented around thinking about the drivers of these themes of value and quality and sentiment.

Celina. Some quantitative investors include growth as an investment theme, but AQE at State Street does not.

Olivia. There’s a negative aspect to growth in our view. So companies that are growing too fast – like say they are growing their assets too fast – that’s actually not a good sign for future earnings. And so we take a contrarian stance against abnormal asset growth in our investment process. We see that as a low quality signal, actually. So, yeah, there are aspects of growth that we are incorporating and then there are some bad growth characteristics we are trying to explicitly avoid.

More Insights

Celina. When we talk about these investment themes, you’ve mentioned that one of the advantages of a quantitative approach is the ability to use data and to apply objectively measurable data to a very wide breadth of companies. How does that actually take shape in practice?

Olivia. I’ll give you an example of the sorts of inputs that go into this. So when it comes to investor sentiment, the types of inputs that we care about are prices of stocks, volumes being traded in stocks, forecasts of earnings and sales by analysts that are researching and publishing their forecasts in the market, hedge fund positioning might be a source of determining what investor sentiment is. So, all of those data inputs are important for investor sentiment. We might even use some data that we’ve obtained about the supply chain and how customers and suppliers are linked in the sense that you can make a forecast about a company based on whether its supplier or its customer is being impacted either in its prices or in its earnings.

From a valuation standpoint, again, there’s many different inputs that go into forming views on whether stocks are reasonably valued or not. But it actually gets quite nuanced when you start thinking about value, because in different industries – like, for example, real estate – it’s the funds flow from operations that really drives the valuation for a for a REIT, or it might be the adjusted funds flow from operations, or it might be the net operating income because you know, rental income is a big driver of the valuation of a REIT. In the case of a healthcare company, the amount of research and development they’re doing is very important in terms of their valuation, and so how we incorporate adjustments into the valuations, based on how much research and development they’re doing is very important. In the case of a bank, the book value is very important, but possibly the enterprise value or the cash flow value is not as important for a bank.

And then from the quality standpoint, quality has just got so many aspects to it. You think about balance sheet or earnings quality. And quality of management can actually also be assessed, funnily enough, with data, in our view. We tend to believe that companies that are focused on the long-term issues that may be affecting their industry or their company directly are in a better financial position to be less focused on quarter-to-quarter earnings and managing shareholder expectations. If they’re able to focus on the long term as well, that is a very great endorsement for a company’s quality. Companies that have strong cash flow as well, that’s a very good sign of quality. That’s just a handful of examples of the types of data. Now bringing all this together involves turning data inputs into what we call signals. And then turning the signals into scores, scores so we can rank companies from best to worst, and then combining all the different scores for all of those signals into the themes and then ultimately into the one company’s stock return forecast.

Celina. So when you cite all of these examples of objective measures that go to each of these investment themes, those are actually statements that are not solely based on your intuition or even your own experience as investors, but actually by a rigorous in-depth research process.

Olivia. Yes. So how we selected these signals in the first place I think warrants a bit of explanation. Did we just have a small group of people that in the early nineties sat down and said, “Right. These are going to be the signals that we use” – and ever since then, we’ve just put our feet up on the desk and cranked the handle and never really thought about it again?

Celina. Because that can be a misconception about what quants really do.

Olivia. Yes, exactly. The process of having a model that is forecasting stock returns every day requires constant evolution and constant evaluation as to whether it is the right thing to have now and into the future. The world is changing on many dimensions, which warrants a review of the way you forecast stock returns. Firstly, businesses change and the composition of the market changes. Secondly, there might be something like a tax change which impacts the way companies measure their earnings or incorporate pensions on their balance sheets, or the way they expense their research and development, the way software changes from being an asset to a service. Those things warrant a revisiting of the way you assess valuation or assess quality or some other thing.

The other thing that can be an opportunity for us is the availability of data has exploded so, if we’re thinking about what investment problems we’re trying to solve, then the data we’ve got available to us to help solve that problem is changing. For example, the supply chain data wasn’t available 20 years ago. The way we assess sentiment for companies is changing as well because of the ability to compute –I guess to do calculations –on vast amounts of data. So, for example, text analysis of financial-statements conference calls. You know what people are saying on social media about companies. Think about how people try and disclose good news from bad news. Think about the language people will typically use when they’ve got good news to give. So, for example if, you're asking me about whether my fund performance is good or bad. If it’s good, I’ll probably say “It’s good.” If it’s bad, I might try and fluff around the language with –you know to try and hide the fact that it's bad. So I might say “Oh, well, you know there was ... there was a lot going on in the quarter, and a lot of volatility and uncertainty, and you know if you check, all things considered, maybe we had a little bit of underperformance.” So, you know, I use all of these words to talk about the fact that there’s a bit of bad news there. The same thing might happen in a company. So, the company management will probably use more complex language in the way they communicate bad news to the market. And so you can actually do analysis on the language and the complexity of language, using computer techniques to analyze that complexity and turn it into a score. I think that’s something that wasn’t available to us 10, 15, 20 years ago, when computing power was just not large enough to process this kind of data, because it is –it’s unstructured data, it takes a lot of processing and it, you know, wasn’t possible to do.

Celina. I’ve heard this described as kind of a fuel for the “Golden Age of Quant,” or a new Golden Age of Quant. But how do you use data well? Data has always been vulnerable to being misused. And it has always been extraordinarily valuable when it is used well. So what have you learned, what can you share about how you use data really well?

Olivia. So I’ll tell you what the danger is first. The danger of misusing data – and the reason why the danger has really increased today. There is no shortage of data availability at low cost, and a desktop computer can do a lot of processing. So the barriers to using data to make investment decisions are lower than ever before. So what this means is, the ability to look for relationships between some of this data and stock returns is so easy – that what’s called “back tests,” – can be done on a whim. The fact that it’s so easy to take some data and run a test between an input data set and stock returns means the likelihood of finding a relationship that is statistical on paper is really high, but the probability of finding something that is spurious and just a statistical fluke or luck is really high as well.

And so the very difficult thing now does is no longer “I need to you know create an algorithm and find the data and put it in my database.” That’s actually not the tough part. The tough part is deciding with great discipline what you’re going to test and what you’re not going to test. And so more than ever before, it’s so important to place a very high weight on outlining your hypothesis and your expectations from an economic and investment intuition standpoint, before any data is looked at, and that is the way we are trying to guard ourselves against finding relationships and data that will not persist in the future. And we call that guarding against data mining.

Celina. And it's big no-no. It's something you really want to avoid doing, because humans are incredibly talented at coming up with stories that will explain any number of completely spurious or chance relationships.

Olivia. Absolutely. And I think when it comes to research, if you’re just thinking it from a pure “I want to discover something that nobody has discovered before,” and the pressure to want to tell the world that you’ve discovered something could lead to this strong desire to rationalize it and publish it in a paper. However, publishing it and writing about it and telling people the idea is not the same as investing in it. We are actually incentivized to be very picky about what we include and the way we forecast stock returns, because ultimately we need to invest our clients’ money in these ideas, and we need to figure out and be disciplined about avoiding ideas that just don’t play out – ideas that sound really good in theory and that could be very exciting for people to be working on. But ultimately the discipline around what you choose to put in the model, you know has to be very high.

Celina. How has this experience – this sort of career-long experience in, on the one hand scrubbing out your own emotional biases as much as possible, and on the other hand considering how to exploit the emotional biases that take shape in the markets – how, if at all, has this changed your own perceptions?

Olivia. So I think there’s two seminal kind of events maybe, that occurred in my career that lead me, you know that have shaped the way I think about investing.So, the first one is really about my very first job in the industry. So I was employed as a trainee equity analyst in Australia and the first project I was asked to do was to research a single company, and it was a casino on the Gold Coast in Australia. I got to know this company pretty well, and I did a very deep dive into its operations and the way it worked. I got to visit the company and that was really exciting for me. I was straight out of university, I got to travel to the Gold Coast, I got to see and meet the CEO of the company, the chief finance officer. I got to ask them, you know, really detailed questions about who was visiting the casino, and where they were coming from, and how much money they were spending and how many people they got through the door. And, you know, I was really on a high about how much I knew about this company. And, you know, I put my model together and at the end of this exercise, I found it near impossible not to love this company and think it was the best idea to buy this company. It was just very difficult to be objective. I’m not going to say we got burnt by this company, and I think ultimately the senior analyst also agreed that it was a purchase. I just distinctly remember feeling like I didn’t want to listen to other perspectives. So, this seeking out information to confirm what I already thought was true, this was confirmation bias. And as my career went on and I started to learn a lot more about behavioral finance and the behavioral aspects of how people make decisions – that experience, you know, coupled with what I had learned, really impacted my desire to have a more objective way of assessing whether companies are worth buying or not.

But I also want to touch on the second moment that was a key learning for me. And this was during the global financial crisis. So during the financial crisis, value was absolutely punished, because in the financial crisis there was a huge flight to quality and it didn’t matter. Everyone was panicked about what was going to still exist in companies and in financial markets after that crisis. And as you’re watching the market plummet, and you’re watching your investments that you’ve invested on clients’ behalf fall even harder than the market – that is very difficult to watch. And you know there was just a lot of panic in every, in every investment market. And just because we have a systematic and disciplined way of picking stocks, doesn’t mean we don't feel this same panic that everybody else is going through. And so the urge to do something and change and make rash decisions during a period of crisis like that took hold of many – just so many institutional investors. The urge to do something that is probably, likely detrimental to your long-term outcome is so strong in periods of stress in markets. And I can tell you in 2008, the urge to increase exposure to high-quality names at the end of 2008 was extremely overwhelming – the urge to change the investment process away from value when it was the exposure you needed the most in 2009. And I felt that and I really learned from that.

Celina. What makes quantitative investing, a uniquely valuable approach to investing?

Olivia. I want to say that there are many different approaches to investing, with many different horizons, and capitalizing on different aspects of what the market offers. So, a quantitative approach, that is grounded in investment intuition, and takes advantage of breadth, can give you diversified portfolios, with limited stock-specific risk, and strong risk control around portfolio outcomes. The risk control comes from being able to compare risk and return in an objective way. So I think that from a core allocation standpoint, a quantitative process with its breadth, with its diversification, with its exposure to the hidden gems, is a very good complement to other portfolio styles that may adopt very high-concentration, very high-conviction positions in individual stocks.

Celina. It’s become incredibly clear to me that you love quantitative investing. What do you love about it, why do you love it?

Olivia. So the thing I love most is our weekly research meeting. This is the moment where the humans come together with the numbers. I mean, that sounds, maybe that sounds weird, but I love the creativity that goes into figuring out how to exploit an investment problem. Using things we can measure and calculate and then implement in an objective way. Just the creativity in investment markets is fun, it’s just really fun. I love the fact that everybody in my investment team has a different experience either in the market they’ve operated in or the educational background they had, or even their career path to date that’s led them to the investment market. I love the fact that they all come at it with just a slightly different angle, or think of things. And I love the fact that I didn’t think of most of the things they thought of. And why I love my job, actually as the CIO, is because I’m responsible for this whole team of people that bring this creativity to the table.

Celina. So it’s a journey of curiosity, of a spirit of discovery. And I thank you for bringing us along in that process.

Olivia. Thank you, Celina I've really enjoyed it.

Celina. To learn more about active quantitative equity at State Street, please visit us at www.SSGA.com/AQE

Glossary

Alpha: Alpha is used in finance as a measure of performance. Alpha, often considered the active return on an investment, gauges the performance of an investment against a market index or benchmark which is considered to represent the market’s movement as a whole. The excess return of an investment relative to the return of a benchmark index is the investment’s alpha.

Backtest: Backtesting is a way to evaluate the effectiveness of a trading strategy by running the strategy against historical data to see how it would have fared.

Behavioral bias: Financial Behavioral Biases are deep-rooted patterns of investor behaviors which, if not managed, can cause a client to make irrational decisions on a regular basis.

Behavioral economics: Behavioral economics, along with the related sub-field, behavioral finance, studies the effects of psychological, social, cognitive, and emotional factors on the economic decisions of individuals and institutions and the consequences for market prices, returns, and the resource allocation. Behavioral economics is primarily concerned with the bounds of rationality of economic agents.

Factors: Factor investing is an investment strategy in which securities are chosen based on attributes that are associated with higher returns.

Factor investing requires investors to take into account an increased level of granularity when choosing securities; specifically, more granular than asset class.

Forecasting: The use of historic data to determine the direction of future trends.

Quality: Quality has long been established as an investment approach, dating back to Benjamin Graham, but it is less well accepted as a factor, especially when compared with value, size, yield, momentum and low volatility. Quality is defined by low debt, stable earnings, consistent asset growth, and strong corporate governance. Investors can identify quality stocks by using common financial metrics like return to equity, debt to equity and earnings variability.

Sentiment: Market sentiment is the feeling or tone of a market, or its crowd psychology, as revealed through the activity and price movement of the securities traded in that market.

Signal: A term used interchangeably with Alpha, is a measure of performance, the excess return of an investment relative to the return of a benchmark index.

Valuation: A valuation is the process of determining the current worth of an asset or company.

Disclosures

The views expressed in this material are the views of Olivia Engel through the period ended November 19, 2018 and are subject to change based on market and other conditions.

This document contains certain statements that may be deemed forward-looking statements. Please note that any such statements are not guarantees of any future performance and actual results or developments may differ materially from those projected.

This podcast is provided for informational purposes only and should not be considered investment advice or an offer for a particular security or securities. The views and opinions expressed by the speaker are those of his or her own as of the date of the recording, and do not necessarily represent the views of State Street or its affiliates. Any such views are subject to change at any time based upon market or other conditions and State Street disclaims any responsibility to update such views. These views should not be relied on as investment advice, and because investment decisions are based on numerous factors, may not be relied on as an indication of trading intent on behalf of State Street. Neither State Street nor the speaker can be held responsible for any direct or incidental loss incurred by applying any of the information offered. Please consult your tax or financial advisor for additional information concerning your specific situation. This video cannot be used for commercial purposes, and should only be used in the United States as restrictions exist with some products and services marketed globally

The whole or any part of this work may not be reproduced, copied or transmitted or any of its contents disclosed to third parties without SSGA’s express written consent.

©2019 State Street Corporation. All Rights Reserved.

State Street Global Advisors Worldwide Entities

2322439.2.1.GBL.RTL

Exp Date 08/31/2020