How To Beat The Market By As Much As The Typical Investor Underperforms

Summary

• In a prior article, I took an inventory of factors that drive risk and return in various investments.
• In today's piece, I will expand on this and explain why major factors exist, structurally.
• You can use the implications to build a better portfolio.

Much of the asset management industry is currently focused on delivering "smart beta" with low fees. The main approach currently in use is to find factors that have historically outperformed the broader market and tilt portfolios towards them. The underlying premise is sound, but the implementations sometimes fail due to poor forecasting ability of the sponsors, transaction costs, or both. However, when used properly, factors of risk and return can be used across asset classes to build very strong portfolios. Let's take a quick look at some common factors in stocks, why they exist, and the groups of investors on the other side.

It's clear that the average retail investor badly underperforms the market. Retail trader performance is so bad that often brokers trade directly against retail by shorting them the assets they buy on margin. A guy named Anton Kreil has some good videos on this (below).

After watching, it seems likely that the HFT firms that buy Robinhood customers' orders follow the same model and simply take the other side of their customers at all times and hedge the positions with long S&P 500 futures. Then, when people lose all their money trading, the funds end up in the accounts of the HFT firms. This is doubly true for options, although the mechanics of the hedges are a little different. Kreil breaks down some conflict of interest patterns for retail brokerages in the video (he focuses on forex and euro-style CFDs, but many of the same conflicts are present in US-facing retail brokerages).

But if what retail traders are doing doesn't work, then what does? Academic finance has some ideas. Anomalies 1-3 are pretty basic but are worth explaining to everyone interested, but longtime readers might want to skip to 4, 5, and 6 if you want to be surprised a little.

1. The value premium

Stocks that have low valuations outperform stocks with high valuations (price to earnings and price to book are the most commonly used metrics). The difference is somewhere in the neighborhood of 1-2 percent per year in the long run, but value has multi-year stretches of underperformance, so it's not a free lunch. Value stocks (via my favorite proxy VYM) have delivered similar risk-adjusted returns to the market over the last 10 years, but have trailed high-flying tech stocks.

Source: Portfolio Visualizer

Who trades against value?

Actively managed mutual funds and retail investors. Actively managed mutual funds have about $12 trillion in assets in the US, as of the last count. This is approximately 4 times the amount of money that so-called smart money hedge funds have under management. Fund managers talk the talk, but in practice, they tend to buy the stocks of large, popular, and expensively valued companies. These companies tend to get the most media attention, and thus the demand for their stocks is the highest, pushing risk premiums (and eventual returns) down. Both retail investors and fund managers tend to gravitate to large-cap stocks, but there are more opportunities in small caps. Even I am guilty of focusing my coverage on popular stocks at the expense of smaller names.

The good news is that value and momentum are natural complements, so if one is doing poorly, the other tends to do well. Combining value and momentum actually doesn't cancel out like combining value and growth, investing equal amounts in value and momentum combine for about a 2-2.5 percent annual edge vs. the market that is more stable than either factor alone.

2. The size premium

As I wrote above, actively managed mutual funds and US households tend to gravitate towards the stocks of larger companies, leaving an advantage (1-2 percent per year) to small companies.

Who trades against the size premium?

Again, actively managed mutual funds and households trade against the size premium, leaving an opportunity to invest in smaller companies for higher returns. The size premium can be a little tricky to exploit, however, because it's correlated with other factors.

There is evidence that the size factor works, however. Size works best when combined with the quality factor, which we'll get to in a few paragraphs (below, IJR is small caps, IVV is the S&P 500).

Source: Portfolio Visualizer

3. The quality premium

The stocks of companies with positive earnings, low levels of leverage, and shareholder-friendly management outperform the companies with the opposite characteristics over time. Quality is a great factor to overlay with other factors, like value and size, because it isn't strongly correlated with the other factors. Again, quality outperforms the market by around 1-2 percent per year on average. I used IJR, which represents the small-cap S&P 600 against the Russell 2000 (IWM) to illustrate.

Source: Portfolio Visualizer

Who trades against quality?

Index funds are the main group of investors that trade against quality. Because they have to buy every stock regardless of how poorly or well it is run, rising AUM in index funds is likely to cause the magnitude of the quality premium to rise over time. The bottom half (in terms of skill) of mutual fund managers also inadvertently buy junk stocks, often under the guise of value investing."Dividend divas" are another group of investors that will buy any stock with a 5+ percent yield, no matter how much of a fait accompli a dividend cut is. They then sell all their stocks for a loss when the inevitable dividend cut occurs.

4. The low-volatility premium

Low-volatility stocks tend to outperform high-volatility stocks, per unit of risk taken. For example, low volatility stocks might have a long-run average annual return of 9 percent per year and risk of 12 percent, and high volatility stocks might have a long-run return of 10 percent per year and volatility of 20 percent.

Here's a quick graph of how this has played out since 2012 (when the world felt the need to invent a high-volatility ETF). As you can see from the X and Y axis, the returns are all about the same, but the volatility difference is huge.

Source: Portfolio Visualizer

Who trades against low volatility?

A cynical explanation for this anomaly would be that fund managers who get paid for performance have a strong incentive to maximize volatility around the time they are due for bonuses or carry payouts. However, the easiest explanation of why the anomaly is so large is that many investors don't have access to leverage and are forbidden from concentrating too much risk in equities, so they are more or less indifferent to taking additional risk that is accepted within their investment frameworks. Such investors appear to be willing to exchange roughly 7 points in volatility per year for 1 point of additional return, however, which is an obvious bias.

Basic portfolio theory would hold that you could buy some low volatility stocks on margin to set the volatility of low vol ETFs equal to the market. As long as interest rates stay reasonably low, you could have made about 3 percent more per year by holding the low volatility asset with 1.25x leverage than holding the S&P. Don't bet your house on it, but it's another solid source of somewhat uncorrelated outperformance. 2018 was a banner year for the low volatility anomaly.

5. The momentum effect

Momentum is one of the strongest anomalies in finance. Stocks that have gone up in the recent past tend to keep rising in the future. Additionally, a small subset of stocks is responsible for much of the overall market return. The easiest proxy for momentum is the NASDAQ 100 (QQQ), which holds the 100 largest stocks listed on the NASDAQ. Contrary to popular belief, momentum isn't inconsistent with other factors like value and low volatility, but rather can be combined in the portfolio in a way that leads to a higher expected return and diversifies risk away from other factors. Very few mutual funds even came close to matching the returns of the NASDAQ 100 since the 2009 low. The NASDAQ exemplifies the momentum approach because stocks are quickly dropped from the index by new companies that overtake them. For those wondering, momentum and low volatility with leverage are the two strongest equity anomalies known.

Momentum historically outperforms the market by between 3 and 5 percent annualized, and the NASDAQ beats the S&P in roughly 60 percent of quarters going back 30+ years.

Source: Portfolio Visualizer

Who trades against momentum?

Actively managed mutual funds and households both trade against momentum due to a bias known as the disposition effect. Stocks that have gone up are disproportionately targets for selling by fund managers, who prefer to hang on to their losers. However, individual winners tend to keep winning and losers tend to keep losing. This bias multiplied by the millions of investors who hold it leads to higher returns for stocks that have recently done well. Investor portfolios that don't know about this anomaly slowly fill up with junk because they sell all their winners to plow more capital into their losers.

6. The leverage effect

The leverage effect isn't as famous, academically speaking, but it is, in my opinion, the number one overlooked key to investing success. The leverage effect basically means this - when stocks go down, investors will be often forced to sell due to the use of leverage in the stock market. Selling begets selling, which leads to large swings from fair value for the market. The flip side is true on the way up, as strong earnings allow companies to aggressively buy back stock and allows investors to borrow more against their newly created mark-to-market equity. Here's some recent academic research diving into this effect for those interested. Volatility targeting is an idea that has gained traction over the last 2-3 years so some pundits are scared of it, but my guess is that the capital imbalances that cause the leverage effect will overwhelm capital on the other side trying to take advantage of it indefinitely. My second guess is that like risk parity, volatility targeting will be mostly ignored in favor of the VC/private equity lottery with massive management fees.

When markets continue to fall, even more liquidations happen and thus markets can fall further still. If you have a long time horizon, however, stocks become a better deal when prices fall, so trading volume tends to rise in times of volatility and market stress. How can you reconcile the fact that markets tend to fall further after declines but become a better deal? Investors have different time horizons so the long-term investors may not care if markets continue to fall over a 1-3 month time horizon if they can get a good price on a five-year time horizon, as was true for those who bought equities down 20-30 percent in 2008 and had to endure further drawdowns. Both groups of investors made money but rationally traded against each other.

Here's the cool part. Let's say you're an index fund investor. If you also invest in a fund that uses volatility targeting or can carry it out on your own, you can make money on both time horizons. Market timers and long-term investors like to argue about who is right, but having a little money in both strategies means you can have it both ways. The skilled market timer can make more accurate forecasts over a 1-3 month time horizon, whereas the buy-and-hold and buy-the-dip investor gets a better price to maximize their five-year expected profit.

To illustrate, here's a simple test I used with a 2x leveraged S&P 500 ETF, with a program to target volatility at 14 percent annualized (a little less than the long-run average for the S&P).

The model beat the S&P by roughly double going back to 2007. If you start it in 2010, the model still wins by over 50 percent. The model wins handily if you start in 2014, also.

Blue is 14 percent volatility targeted S&P 500, red is 2x leverage, and yellow is 1x leverage.

Source: Portfolio Visualizer

The flip side of the leverage effect is to use volatility targeting in one form or another to complement your equity investments. It requires leverage to execute, much like the low volatility anomaly but can add about 2 percent to your returns per year.

My preferred way to use volatility targeting is to use a slightly modified (to filter out noise) 200-day moving average for equities. I find that it's a purer way to exploit the leverage effect since the 200-day moving average is a good guess of where the average margin investor purchased their stock. Margin tends to build and build during strong markets, only to be washed out in bear markets. You can profit from this.

Here's a (somewhat cynical) graph from the Wolf Street blog. I might be mistaken, but I think he has contributed to Seeking Alpha as well. It shows the leverage building in the system, which clearly has grown faster than the market capitalization of the S&P 500.

Source: Wolf Street

Who trades against the leverage effect?

Mostly retail traders, but also a lot of hedge funds and other aggressive traders.

Investors who borrow money from their brokers to invest trade against the leverage effect. Collectively, investors have hundreds of billions of dollars in margin debt that can be called at any day the market threatens the collateral of investment banks that make broker-dealer call loans. Investors are liable to be forced to sell their stocks when the market drops.

I think the best way to take advantage of the leverage effect is to use the 200-day moving average to trade S&P 500 and Nasdaq futures on a 1-3 month time frame. Some hedge funds like to use pure volatility targeting, but I'm not as big of a fan because it has higher transaction costs.

The trick is to get out before other investors are forced to and then to buy back stocks lower on the following upswing.

A 60-second implementation of factors 1-6 might look like this:

1. 20 percent S&P 500 futures (use the 200-day moving average and switch to Treasury futures when the S&P 500 is below the 200-day MA)

2. 30 percent NASDAQ futures ("^")

3. 35 percent VYM

4. 50 percent USMV

5. 25 percent cash

6. Rebalance monthly

This is for illustration purposes only, because there's no good reason to build a standalone portfolio in only one asset class. The blue graph is our portfolio, and the yellow is if you only used the leveraged ETFs and went all in on the leverage effect. Green is the S&P 500.

The backtest goes to 2012. If I start it before 2008, it's extremely favorable to the timing model and unfavorable to buy and hold. The youngest ETF in the test started in 2012, so that was the date chosen.

Source: Portfolio Visualizer

The portfolio had its hands tied behind its back due to using high cost leveraged ETFs used for the test but still managed to beat the S&P by 3.9+ percent annualized (I estimate the outperformance for using futures as a small fund at about 50-75 bps more than the data shows for LETFs, with a corresponding decrease in NAV volatility).

If you start tests for volatility targeted portfolios before 2008 their advantage tends to be a little overstated due to avoiding the bear market of 2008, and if you start the tests after, the results tend to overstate the advantage of buy and hold relative to the long-run average by not counting the bear market. The Sharpe ratio of the factor portfolio (blue line) was 1.46 over the period tested vs. 1.27 for the market (both numbers are historically high). Where you start to have fun is when you make an equity factor model like I've done here and throw it into a risk parity portfolio, after you've made factor models for all the other asset classes. Then you can start to get some more consistent returns!

Note that the additional return from adding factors isn't the sum of all the edges because a lot of them are correlated with one another. However, using factors has been shown to provide a sustainable advantage over the broader market. Also notable is that the S&P 500 is better than it looks at first glance because it was designed to take advantage of two factors, namely quality, and momentum. That's part of what makes the S&P so hard to beat for mutual funds. The median stock returns about 5 percent per year, so if mutual fund managers aren't wise about factor investing and momentum/disposition bias, they're swimming upstream from the start.

Conclusion

For stocks, you can combine all of these factors to get a return that's about 4.5 percent higher per year than the long-run average of about 9-9.5 percent. The tricky part of implementing the ideas is to understand the correlations and understand how to size your bets.

I don't think it's an accident that well-designed factor portfolios seem to outperform the market (~+4.5 percent annually) by about the same margin that retail fund investors underperform (-4.5 percent annually). By stringing together a few relatively weak forecasts, we can make much stronger forecasts that can produce more consistent profits. You can do a lot better by applying this same logic across 5-7 asset classes rather than 1 asset class.

I've only covered stocks in this article, but I have similar factor models for everything from Treasury bonds to currencies, credit, commodities, international vs. US stocks, mortgage-backed securities, and real estate. The goal I'm working towards is to find enough edges to get a Sharpe ratio of 2.0, meaning I almost always make money with the average return centered on about 20 percent annualized and risk of 9-10 percent. Here's a semi-recent recap of these, if you're interested.

Lastly, mainstream academic finance is great at telling you that there's a 12 percent chance of snowfall on any given day in Chicago, but knowing whether it's summer or winter makes all the difference in the world. Retail investors, however, are clearly are betting on snow in July.

Disclosure: I am/we are long QQQ, SPY, VYM. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.