Long before Y2K, there were groups of investors that rejected fundamental analysis as a method of picking stocks. They were a mysterious group operating at the outer fringes of the market: totally legal but capable of creating sudden and dramatic price changes. Most commonly dubbed “program traders” or “soes bandits”, these folks created massive order flow using computer programs.
With the exception of operating with the blessing of the New York Stock Exchange, Nasdaq, and other rule-makers, there was very little difference between program trading and what the Securities & Exchange Commission called “pump and dump” schemes.
Techniques for measuring the impact of program trading were largely lacking back them. But one thing was evident. Program trading yielded large and rapid profits for those skilled in the practice and with lots of capital. To be quite honest, this claim may not be verified with absolute certainty. However, the fact that the technique was quickly replicated by others giving rise to an explosion of program trading is evidence enough. Invariably, capital is drawn to the source of excess profit.
Program Trading Morphs Into Quantitative Investing
Before going much further into just how this trading and investment phenomenon became mainstream, let’s first define exactly what we are talking about. Program or quantitative trading consists of trading strategies based on quantitative analysis. This means creating and relying on mathematical computations and number crunching to identify trading opportunities.
Quantitative analysis (QA) is a technique that seeks to understand behavior by using mathematical and statistical modeling, measurement, and research. The purpose of quantitative analysis digest a large number of variables, such as asset prices and trading volumes but also how real-world events may affect asset prices.
These days quantitative trading is widely used by financial institutions and hedge funds. Transactions are usually large and involve the purchase and sale of hundreds of thousands of shares and other securities.
Practicing The Techniques
Quantitative techniques provide analysts with tools to examine and analyze past, current, and anticipated future events. Any subject involving numbers can be quantified; therefore, there are many fields in which quantitative analysis is used and is beneficial. For example, quantitative analysis is used in analytical chemistry, financial analysis, social science, and organized sports. In the financial world, those who were once referred to as program traders are now referred to as “quants” or “quant jockeys.”
But the big guys aren’t the only ones involved. Individual investors are increasingly using quantitative trading techniques. Applications like QuantStart and others offer pre-packaged programs for strategy identification, backtesting, execution, and risk management.
And quant applications offer great value beyond Wall Street. Governments rely on quantitative analysis to make monetary and other economic policy decisions. Governments and central banks commonly track and evaluate statistical data such as GDP and employment figures.
How Big Is Quantitative Investing
For all the advancements in the ability for computers to analyze enormous quantities of data, the extent of Quantitative investing remains open to estimates and interpretation. According to Alex Foster, author of “The Edge of Foresight” and VP at Quantiacs, here are some data points.
Roundly, 90% of volume in the public markets in the United States is traded by quantitative means. This methodology is spreading at rates above average at 10.3% according to the report Global Algorithmic Trading Market 2016–2020. This same report claims that Quantitative Finance is a $1 trillion market.
The report further highlights how 6 of the top 10 performing hedge funds, as of earlier this year, were quant funds. As we go through our discussion of different quant investment styles later in this article, it is significant that most of the top-performing quant hedge funds specialize in commodities.
Semi Quant: Mixing Quant With Fundamental Analysis
Pure quantitative analysis has proven itself a useful evaluation tool. It is becoming more common for investment funds to weave in other factors that are more difficult to measure. This is where Qualitative Analysis steps in focusing on meanings that involve sensitivity to context rather than the singular desire to obtain universal generalizations. Qualitative analysis works by establishing rich descriptions rather than quantifiable metrics. The qualitative analysis seeks to answer the “why” and “how” of human behavior.
“Qualitative analysis works by establishing rich descriptions rather than quantifiable metrics. The qualitative analysis seeks to answer the “why” and “how” of human behavior.”
In a combined qualitative and quantitative analysis project, a company, analyst, or investor might wish to evaluate the strength of a product. Qualitative tools used for the project can include customer surveys and panel discussions. A quantitative analysis of the product can also be initiated through the examination of data regarding numbers of repeat customers, customer complaints, and the number of warranty claims over a given period.
Quantitative analysis is not the opposite of qualitative analysis; they are just different philosophies. Used together, they can provide useful information to make informed decisions that promote a better society, improve financial positions, and enhance business operations.
Taking Away The Emotion
If emotion is truly the enemy of rational decision-making, then Quantitative Investing is the answer. Be it fear or greed or just becoming overwhelmed by mountains of data, emotions serve only to stifle rational thinking, and that usually leads to losses. Quantitative trading does not have these problems. Computers and mathematics do not possess emotions; so quantitative trading entirely eliminates this problem. Computers can be used to automate the monitoring, analyzing, and trading decisions.
“Be it fear or greed or just becoming overwhelmed by mountains of data, emotions serve only to stifle rational thinking and that usually leads to losses. Quantitative trading does not have these problems.”
But while all of this is good, quant investing is not the perfect solution for everyone. Financial markets are some of the most dynamic entities that exist. Therefore, quantitative trading models must be as dynamic to be consistently successful. Many quantitative traders develop models that are temporarily profitable for the market condition for which they were developed, but they ultimately fail when market conditions change.
Some Examples Of How It Works
There is no single quant strategy for all occasions. As long-term market changes have taken place strategies fall in and out of favor. Below are six different categories where quant strategies can be identified.
Systematic Trend Following // CTA
If you recall from earlier in this article, we mentioned how 6 of the top 10 best performing quant hedge funds in 2017 specialized in commodities. This leads naturally to starting with CTA, which is an acronym for Commodity Trading Advisor. A CTA acts much like any other financial advisor, except limited to providing advice related to commodities trading. It is worth noting that many CTAs have moved away from just commodity trading and now invest across all the tradable asset classes.
CTA and the strategy of System Trend Following become almost interchangeable in recent years. This technique is heavily dependent on price volatility that triggers a predetermined price direction in either direction. Once this event occurs a position can be taken. For example, this could include futures in stocks, fixed income, currencies, or commodities such as energy.
Practitioners of this approach are typically high volume traders withholding periods limited to a few days or weeks. If this sounds like little more than day trading, you could be right. Between 60%-70% of these types of trades are losers. However, the rest tend to pay off big time.
As the name implies, Statistical Arbitrage deals with numbers. These days this spells big data where high-speed computers and the closer those computers are to the source of data the better.
SA seeks mispricing by other market participants by identifying relationships between securities where computers sort out anomalies, then placing a bet that normal pricing will be restored. The arbitrage usually is accomplished by investing both long and short two different securities or groups of securities. For example, a fund might belong to General Motors and Ford but short Tesla.
Much like CTA, holding periods can last for a few seconds to as long as a week or more. Either way, this requires big winning bets to offset a high percentage of losers.
If your investment temperament is geared more to longer-term holdings then Factor Investing might be worth considering. In this case, the holding period could run over multiple years.
Here quants identify certain factors that make up a successful company or a successful stock. For example, a search is made for each of the quantifiable factors that go into the success of Apple, Amazon, or Google. Then the manager goes about investing in a position in each of the companies that share those characteristics.
Unlike CTA which seeks to maximize Alpha, Factor Investing is considered more of a risk management (beta) approach. It takes advantage of behavioral biases and mistakes that investors reliably make. For example, people tend to undervalue less-glamorous stocks. Factors are also sources of risk that reward investors with superior returns over time.
Another strategy that typically requires longer-term horizons is the Risk Parity strategy. This is the type of asset-allocation strategy that aims to hold an equal amount of risk among investment classes, which react differently to market changes. The trader diversifies — among, say, fixed income, equities, and inflation-risk assets — based not on price but on volatility or some other measure of risk. The less volatile an asset, the bigger weight it gets in the portfolio.
Though the typical holding period can run into years, adjustments can be made along the way.
For example, if the result is a bond-heavy portfolio with too little upside, leverage is often used to make bigger bets. A common practice at the end of each month is to calculate the volatility and rebalance.
This risk-mitigating strategy aims to produce a smoother ride, with diversification helping the fund during difficult market stretches.
Systematic Global Macro
Broader and somewhat more patient than CTA, as holding periods can include weeks, months, or even years. This approach trades across asset classes and countries and relies upon macroeconomic principles. Using data such as inflation, unemployment, and consumer spending, the strategy attempts to build a set of rules that govern the relationship between economic cycles and market movements.
Here is a more or less common attempt to capture the spread between different currency rates. First, sell low-interest-rate currencies. Offset this by buying higher-interest-rate assets. This is an example of what is called a “carry trade”.
Funds that employ this strategy benefit from diversification across asset classes. It’s often billed as a risk-mitigating strategy, participating on the upside but protected on the downside.
Event Driven Arbitrage
If you don’t mind high portfolio turnover but want to improve on the low batting average of CTA then Event-Driven Arbitrage provides an option. This one reflects a classic hedge fund strategy, anticipating corporate actions and events, with an algorithmic approach. It exploits mispricings that occur before or after analyst revisions, share buybacks, bankruptcies, and the like.
Quant Investing Hits A Rough Patch
A fundamental rule of economics goes something like this. When supply overwhelms demand, prices decline. What this means for qualitative investors is similar. Over time the supply of investor capital allocated to fund managers that practice quantitative investing techniques has risen from just a small share to approximately 90% of today’s trading volume. This makes the odds of anyone or a group of managers achieving above-average returns exceedingly difficult.
According to a recent CNBC report, fewer than 40% of institutional managers outperform their respective benchmarks. The figure can be somewhat misleading since it includes all managers whether or not they are quantitatively driven. However, with 90% of trading tagged to quant strategies, the margin of error is enough to state the point that in the overall sense, the majority of quant managers are no longer providing widely superior returns for their clients. Recent results for 2018 seem to confirm this conclusion.
According to data from HFR, during the first half of this year, quant equity funds lost 1%, while quant macro funds fell 4.2%. This marked the worst performance in eight years; clearly a disappointment. Only 17 percent of large-cap active quant mutual funds outperformed the Russell 1000 index in June, the worst monthly showing this decade.
According to Bank of America Corp. The average quant fund tracked by The Bank of America gained just 0.1% versus 0.6 percent for the broader gauge. Vanguard Group’s $1.6 billion market-neutral mutual funds fell 1.8 percent, the worst in more than a year. Not every strategy is equal. Value has been a poor performer for years now, but recently other popular strategies, such as momentum and quality have also spluttered. Even the highly respected commodity-based CTA strategies are struggling. Through the first 10 months of this year, the Barclay CTA Index is down 2.85%.
What is happening?
There are several possible explanations. First, factors are often called risk premiums because they represent the extra compensation investors, in theory, get for accepting some specific risk, such as the greater volatility of cheaper “value” stocks. Sometimes the factors simply fizzle, and other times the risks manifest themselves in a dismal performance.
So quants that have previously relied on stronger performing factors to counteract weaker ones have been wrong. Second, factors may also be suffering from “ crowding”, as money has gushed into various quantitative investment strategies in recent years. That might help boost gains in the short run but ultimately erodes them. Just like any individual stock or sector, factors can from time to time become overvalued and suffer fallow periods.
Interest Rates: The Pendulum Has Shifted
Since 1982 the most powerful force in the financial universe has shaped virtually every market. That force is falling interest rates. Rates are now moving in the opposite direction, with the 10 Year Treasuries recently climbing to nearly 3% for the first time in a decade. That is wrecking any model based on economic variables in the post-1980’s era.
Little wonder how a Deutsche Bank analyst recently discovered that many of their low-volatility, momentum, and dividend yield funds had “sizeable negative exposures” to rising interest rates.
Most trading strategies are based on or tested against historical financial data. Next, they are then given an “out of sample” test to see how they do in live markets. If they hold up under fire, quants design trading algorithms to systematically implement them.
But these backtest have at least one major weakness: the further back in time you test, the less detailed the data becomes. Data that goes back further than the 1980s is almost useless. In other words, the only really useful data comes during a one-direction market for interest rates. The pendulum change in rates provides some insights but doesn’t entirely answer the question. Many quant funds did exceptionally well in the years leading up to the financial crisis. That was the last time interest rates were rising.
So for now the best explanation is probably a mix of crowding and the paradigm shift in interest. No strategy can work all the time. Quants are constantly tinkering with their models to avoid long stretches of underperformance when the market regime shifts.
Newer Is Not Necessarily Better
These newer more sophisticated quant funds go well beyond mining the big factors that can be packaged into simple ETFs. However, the quant factories producing the fund managers of tomorrow have also been hurt by their underperformance. The average pairwise correlation of global stocks, essentially a measure of how closely they move together, has bounced sharply from the multi-decade low touched this year. This only confirms the cause of the recent quant under-performance could be deeply rooted in a design that has yet to adjust to the new reality.
So investors who have fallen in love with quant funds irrespective of the particular strategy employed, the duration of the holding period, or the implicit risks still need to keep an eye on seismic changes in the interest-rate environment.