Beat the Market (Part 3) - Building Blocks of a Market-Beating Systematic Strategy

Why superior? Because it has delivered a higher long term return with a much lower drawdown.

Why the SPY?

The S&P 500 index is the most popular stock benchmark out there and the one to "beat" for any aspiring active investor. Of course, we can't directly invest in an index and so the best investable alternative is the SPY, which closely tracks the index (minus a small fee). As mentioned in Part 1, you can use the VOO and IVV ETFs if you wish, but we're using the SPY as it has a longer history for our backtesting purposes.

Why use a benchmark?

Because if you can't beat the SPY, why waste all the time and effort needed to actively invest? Just buy the market ETF, don't look at stock charts too often (it'll mess with your emotions), and go enjoy the rest of your life.

The benchmark is the investment hurdle that all active investors have to beat to justify their efforts. If they can't beat the benchmark in the long term, then they should just adopt a passive strategy of "Buying and Holding the Index"¹.

What is the "Beat the Market" strategy?

In a nutshell, it's a monthly ETF rotation strategy.

"Monthly" means that we only need to evaluate the strategy once a month, to see if any trades need to be made. Specifically, we will run the strategy at the end of every month, and if trades need to be made they will be executed at the closing price of the last trading day.

Why monthly? Why not daily, weekly, quarterly or yearly? Those are all possible time periods to use, but I've chosen monthly as a balance between ease of use (daily and weekly will require more frequent adjustments and higher turnover), and the adaptability of the strategy to market conditions (quarterly and yearly are less responsive to market movements). I believe most investors will find spending 10 minutes a month to run the strategy and execute the trades a worthwhile use of their time to get a higher return with lower risk.

A "rotation strategy" means we own one ETF at a time, and if the rules of the strategy indicate we should switch to another one, we will sell the existing one and buy the new one, i.e. we will rotate from one ETF to another.

Which ETFs are we using?

In the basic form of this strategy (we'll be enhancing it in the next part of this series) we'll be using the following ETFs:

SPY (SPDR S&P 500 Trust) – represents a broad-based exposure to the US stock market's largest companies.

EFA (iShares MSCI EAFE ETF) – represents a broad-based exposure to non-US stock markets globally. EAFE stands for "Europe, Australasia, Far East" and includes large and mid cap stocks from across 21 Developed Markets countries² around the world, excluding the US and Canada.

QQQ (Invesco QQQ Trust Series 1) – represents an "innovative company" exposure and tracks the 100 largest non-financial companies listed on the Nasdaq by market cap. It's a tech-heavy and growthy index that has captured (and will hopefully capture) many big structural trends in the future.

SHY (iShares 1-3 Year Treasury Bond ETF) – a cash-like safety asset that holds short term Treasuries.

What are the rules of the "Beat the Market" strategy?

The basic version of the "Beat the Market" strategy is very simple:

1. At the end of every month, look at the total return (i.e. including dividends, if any) of SPY, EFA, QQQ and SHY for the previous 3 months (specifically, we use the previous 63 trading days to be consistent, with each month having roughly 21 trading days on average)³
2. Buy the ETF with the highest return of this group if you don't currently own it (while selling the currently owned one), or keep holding it if you do⁴.

That's it!

It's a relative momentum strategy as it picks the ETF with the highest relative strength, and also uses absolute momentum as SHY is a cash-like ETF that has returned ~2% over the long term (which then becomes the hurdle rate that the other ETFs have to cross to be chosen).

A note about the 3 month lookback period

Why use a 3 month lookback period (time period which we use to measure and compare returns)?

Most researchers find that lookback periods of somewhere between 3 to 12 months work well for momentum strategies. Which lookback period works the best is a controversial topic.

Academic research tends to find that over long periods of time a lookback period of 12 months works the best. But this can vary over different time periods and markets.

In my testing, a 3 month lookback worked best from a risk and return perspective, but all time periods from 3 to 12 months matched or outperformed the benchmark SPY from both a Compounded Annual Return and Maximum Drawdown perspective.

Using shorter lookback periods means the strategy will react quicker to changes in trends, but will also be more susceptible to "whipsaws" – exiting a position after a short period of underperformance, only to go back in again after missing the initial period of recovery.

Using longer lookback periods helps to reduce turnover but also means the strategy will react slower to trend changes.

Personally, I chose 3 months because it had the best results and I prefer a strategy that reacts quicker to trends. I don't mind the higher turnover as transaction costs are very low (trading commissions are low nowadays and the ETFs are highly liquid), but your situation (e.g. taxes, investable capital etc.) could be different so you can choose a different lookback period if you wish (a 12 month lookback, i.e. 252 trading days, works well too).

As mentioned above, any choice of lookback period between 3 and 12 months would have given you a similar or higher return and lower maximum drawdown versus the SPY benchmark, i.e. a better risk-adjusted return.

What have the returns for "Beat the Market" been historically?

It's a really simple strategy, but the returns are impressive.

NOTE: This is not financial advice. Results are hypothetical, do not indicate future results, and do not represent returns any investor actually attained. Please read the Disclaimer page for more information.

At the end of the 18 year period from 2003 to 2020, you've compounded at ~15% (vs. 10.3% for the SPY) and ended up with more than twice the money.

You've also done it with a lot less volatility and pain, with a Maximum Drawdown of just 22.5% vs. 55.2% for the SPY.

There is a cost – you'll be making 82 trades over this 18 year period or around 4.6 trades per year on average (the range is from 2 to 7 trades a year). While it sounds scary that your entire portfolio is turning over 4.6 times per year on average, the slippage should be minimal (unless your portfolio is in the tens of millions of dollars and above) as these are very liquid ETFs.

Also, there will be years where you'll underperform the SPY. In fact, you'll be underperforming the SPY for 8 out of those 18 years, by up to 15%.

While the long term performance pick up of the Basic "Beat the Market" strategy is impressive, to me the main advantage is the much smaller amount of downside it takes to get it. This will help you sleep better at night and stick with the strategy.

Imagine a year like 2008 during the Global Financial Crisis, when markets around the world collapsed and the SPY ended the year down 36.8%. With the Basic "Beat the Market" strategy you were actually up 0.5%!

It gets better – in the next part of this series we will look at some enhancements to further improve the risk-adjusted return of the "Beat the Market" strategy.

Footnotes

1. Actually, the "Beat the Market" strategy invests in ETFs based on different indices and also bonds, so a more appropriate benchmark will be a "Buy and Hold" allocation to each of these (rather than just 100% in the SPY), but we'll deal with that at a later time.
2. Developed Markets countries in the MSCI EAFE Index include Australia, Austria, Belgium, Denmark, Finland, France, Germany, Hong Kong, Ireland, Israel, Italy, Japan, the Netherlands, New Zealand, Norway, Portugal, Singapore, Spain, Sweden, Switzerland and the UK.
3. To prevent a "look forward" bias, we look at the total return over the previous 63 trading days excluding the last day (e.g. on 31 March we look at the period from 30 March and count 63 trading days back from that date).
4. In our backtests we use the closing price of the ETF on the last trading day of the month, which means that in real life you'd have an entire day to place the order.