Executive Summary

For decades, the universe of institutional investors has been largely split into two camps: fundamental investors and quantitative investors (quants). The former generally made investment decisions based on earnings statements and, of course, company fundamentals. The latter were less concerned with the CEO’s quarterly comments and more interested in what historical data could tell them about the price of the security going forward. Today, however, we are in a world in which long-understood definitions are no longer clear, and the lines defining investors, dealers, market makers, and investment analysis approaches are increasingly blurry.

Fundamental managers are using machine learning to analyze earnings statements and examine market moves based on past company or macroeconomic news, an approach sometimes referred to as a quantitative overlay. Conversely, quantitative analysts now have access to alternative data sources—such as store traffic from mobile phones or social media sentiment—that allows them to systematically make investment decisions on individual names. Furthermore, the majority of the world’s leading asset managers have both fundamental and quantitative funds. As the lines continue to blur, a new investment approach has emerged—quantimental. These evolutions make the once easy task of defining quantitative investing all the more complex.

For decades, the universe of institutional investors has been largely split into two camps: fundamental investors and quantitative investors (quants). The former generally made investment decisions based on earnings statements and, of course, company fundamentals. The latter were less concerned with the CEO’s quarterly comments and more interested in what historical data could tell them about the price of the security going forward. Today, however, we are in a world in which long-understood definitions are no longer clear, and the lines defining investors, dealers, market makers, and investment analysis approaches are increasingly blurry.

Fundamental managers are using machine learning to analyze earnings statements and examine market moves based on past company or macroeconomic news, an approach sometimes referred to as a quantitative overlay. Conversely, quantitative analysts now have access to alternative data sources—such as store traffic from mobile phones or social media sentiment—that allows them to systematically make investment decisions on individual names. Furthermore, the majority of the world’s leading asset managers have both fundamental and quantitative funds. As the lines continue to blur, a new investment approach has emerged—quantimental. These evolutions make the once easy task of defining quantitative investing all the more complex.

Defining the Market

U.S. institutional investors, which include pensions, endowments, foundations, and unions, currently manage $12.4 trillion in assets, according to Greenwich Associates research. Of that, they allocate just over 4% to hedge funds—or roughly $520 billion.

While not all hedge funds are quantitative and not all asset managers are fundamental, we believe this is a good proxy for quantitatively managed assets in the U.S. Including quantitatively invested assets in the rest of the world would push the number above $1 trillion. Well-known hedge funds and market makers that would fit into this category include AQR, Citadel, DE Shaw, Point72, Renaissance Technologies, and Two Sigma.

From the sell-side point of view, quantitative investing has become a large piece of their revenue. For instance, U.S. equity commissions paid by investors in 2017 total just shy of $8 billion. Forty-one percent of that revenue comes from hedge funds, or roughly $3.3 billion, up from 36% in 2014. Exchanges are in a similar situation, with quantitative investors and others with more automated trading strategies generating at least half of the market’s volume on any given day.

Of course, no two quantitative investors are created equally. Their approaches, focus and styles can vary greatly: equity, fixed income, macro, momentum, and relative value, just to name a few. That said, there are a number of similarities that span the universe and help define the segment. Generally, the quantitative investing process requires research to determine what to invest in, followed by trading signals and execution strategies that optimize the timing of buying and selling securities.

Some of the critical factors defining a quantitative fund are:

Price Prediction

First and foremost, quantitative strategies are tasked with predicting the fair value of an asset at some point in the future. Forecasting future prices or relationships based only on information available in the present is at the heart of quantitative (and arguably all) investment strategy. The methodologies employed and data utilized vary widely, but the end goal is the same—predict the future value with enough accuracy to profit from the strategy, net of trading costs. Forecasting models require substantial back-testing across multiple scenarios to fully understand how well they can predict the future.

Time Horizon

Quants have shorter time horizons and holding periods than fundamental investors. On one extreme, high-speed market makers hold the products they are trading for no more than a day and can have a time horizon measured in minutes or even less. Conversely, hedge funds might take a view over a month or a quarter, hoping to profit from a perceived anomaly in the macro data for which they’ve taken a directional bet on a particular sector or name. As time horizons pass beyond three to six months, quantitative investing drops dramatically.

Holding Period/Turnover

Short time horizons lead to a much higher turnover of total assets in quant funds. As the factors in the models driving investment decisions are recalculated, the portfolio’s holdings must be rebalanced. Assuming such rebalancing turns over 5–10% of the portfolio each month, quantitative funds on average turn over 80–120% of their assets under management annually. The turnover ratio in quant funds can be considerably higher, with some high-speed strategies turning over the entire portfolio monthly or more.

Buy-and-hold mutual funds have turnover ratios of 20–30% on average. And while growth mutual funds that follow a more fundamental strategy can turn over significantly as well—approaching 100% in some cases—managing such a high velocity portfolio today can only be done effectively using automated investing and trading methods.

Transaction Cost

Clearly, shorter holding periods and/or higher turnover can lead to higher transactions costs. As transaction costs can quickly reduce alpha, quantitative managers need to be very conscious of these costs and include those estimates into their trading models. Generally speaking, strategies that are net “takers” of liquidity need to be more conscious of execution costs than those “makers” of liquidity. While the latter profits from capturing the bid-ask spread and collecting exchange rebates, the former is frequently forced to pay exchange fees to access the liquidity their strategy needs.

Automation

The investing process requires research to determine what to invest in, and then trading to buy or sell the agreed-upon instrument. Both processes are separate and distinct and can run the gamut from completely manual to fully automated. Some asset managers might utilize technology to help with portfolio construction but then execute over the phone, while others might rely on fundamental analysis to pick stocks but use execution algorithms to implement the strategy.

Quant funds, however, automate both the research and execution processes. For example, hedge funds electronically trade 46% of their U.S. equity trades on a commission-weighted basis, compared to longonly managers who e-trade only 39%.

It is important to note that some quant investors weight the importance of one over the other. For instance, speed of execution is particularly critical for market makers, whereas deep and fast research matters more for macro funds with a time horizon of a month or more. Nevertheless, automation of both at some level is a critical element of all quantitative investors.

The Impact of Quantitative Investing

Understanding traditional stock picking is relatively easy for most. Will the company grow or not? Will the new iPhone be a big hit? The answers to those and similar questions lead to the decisions to buy or sell. Quantitative investing isn’t so straightforward, with decisions often based on data and mathematical models that very few without a PhD can quickly understand. As a result, quantitative investing has been demonized unnecessarily over the past decade. The proliferation of “flash crashes” has highlighted the changing nature of the markets, with many concluding that automated trading—particularly automated trading by quant funds—was to blame for this bad volatility.

While the speed of investing and trading can certainly exacerbate market moves, it is merely a consequence of market-wide innovation. In fact, in addition to the revenues quants generate for broker-dealers and exchanges alike, the benefits they bring to the market as a whole far outweigh any actual or perceived negatives.

Liquidity and Arbitrage

Following the credit crisis a decade ago, the largest broker-dealers were forced to rethink their business models in light of new regulations and a changing market structure. As a result, they pulled back their liquidity provision broadly. Case in point, agency execution now accounts for 38% of commission spending in U.S. equities compared to 28% only two years ago in 2016, demonstrating the reduction in risk-taking by top brokerdealers. This left a well-documented liquidity hole for investors, who have since looked to new trading venues and new trading counterparties to fill the gap.

While the economic incentives of quant investors are very different from broker-dealers, the turnover they generate acts as a different, yet nearly as important, source of market liquidity. If all the quant funds in the world stopped trading tomorrow, broker-dealers and long-only investors alike would be scrambling to get orders filled.

This rings even more true for those firms whose primary strategies revolve around market making. More often principal trading firms (PTFs) rather than hedge funds, make money by capturing the bid-ask spread millions of time a day. In some markets and securities, they have effectively taken the place broker-dealers filled in the late 1990s before Reg ATS, decimalization, and Reg NMS made manual market making obsolete.

In a similar vein, PTFs also help to keep price correlations in check. Whether that be the price of Coca Cola versus Pepsi, S&P 500 ETFs versus the full basket of 10-year U.S. Treasurys and the corresponding futures, quantitative firms in constant search of slight price abnormalities keep the market efficient and honest.

Disagreements are Good

Possibly the biggest benefit of having quantitative investors in the market is the diversity they bring. There are only so many views fundamental investors can take on the economy, Apple stock or the direction of interest rates. Price targets might differ but, by and large, sentiment at a macro level is often aligned firm to firm. Without an opposing view—which is often the role of the quant investor—the market will lean in the same direction, limiting sellers or buyers when either is desperately needed.

This role of the opposing viewpoint can play out in two ways. First, during normal market times, it keeps the market from swinging too wildly and ensures liquidity in instruments even when the broader market sentiment is aligned. In volatile times, quant investors can act as the shock absorber, a role banks no longer can or want to play.

Prices will likely still drop dramatically in times of stress; however, quantitative models can signal when prices are so low that the profit potential is worth the risk. This is what happened during the financial crisis in 2008, as Lehman Brothers’ debt collapsed, and most investors were looking to unload their “toxic” assets. Quant investors are not in the business of catching the falling knife, but they do certainly look for bargains when everyone else is running for the hills.

The Future of Quant Investing

Quantitative methods will only grow in popularity in the coming years. An influx of new data sources and tools that allow almost anyone to scour even the most unstructured data sets in search of trading signals will continue to attract both assets and quantitative analysts to manage them. It’s not simply about finding new insights from existing data, but finding new insights from new data that, in many cases, didn’t even exist a decade ago. Who needs same-store sales when you can instead view the exact number of shopping bags leaving a store in real time?

To that point, quantitative investing isn’t just for elite hedge funds either. As mentioned earlier, fundamental managers will increasingly employ quant overlays to help inform their securities selection process. Even passively invested index-tracking funds will use quantitatively driven technology to rebalance portfolios, manage createredeems and determine which subset of a given index to hold while limiting basis risk. And let’s not forget about robo-advisors—a crystal clear use of quantitative modeling applied to passive investment vehicles.

In addition to the influx of data, conversations with quantitative investors and experts for this research revealed machine learning as one of the biggest drivers of quant investing in the future. Models that can truly learn from their actions and the reactions of the markets without any human intervention could inject a whole new category of investing strategies. As with all innovations related to automated trading in the past two decades, risk controls surrounding such strategies are key to their success. One rogue machine would sour the market on their utility and safety for years. However, the opportunities presented to investors by machine-learning techniques cannot be understated.

Ultimately, quantitative tools of today and tomorrow present opportunities for nearly all market participants—not only for sophisticated hedge funds with teams of physicists and data scientists. The goal is not to eliminate humans from the process but to provide them with better and more actionable information than was possible with more traditional investment research tools. After all, markets are notoriously irrational, and investing in an irrational market is a team effort. It’s not man versus machine, but man with machine.

Methodology

Throughout March and April of 2018, Greenwich Associates spoke with experts in quantitative investing to better understand their role in the marketplace, broadly held misconceptions about quantitative investing and how tools used by quantitative investors are benefiting the broader investment community. This research also leveraged interviews with 262 U.S. equity investors in the first quarter of 2018 and 472 institutional investors (i.e., pensions, endowments) throughout 2017.