In 1983 two experienced Chicago traders decided to settle a bet: is trading an innate skill or something that could be taught?
To settle this bet, they took out an ad and attracted ordinary people they dubbed the "turtles" and put them through a boot camp to teach them a trend following system.
The Turtles are now legends after they amassed over $175 million over the next 4 years.
Their story shows that trading can be taught and the power of trend following systems. One little, not-so-insignificant detail that often gets overlooked is that the Turtles were given $1 million each to start trading.
Yes, the results are impressive, but there's a reason they were given so much money: they were trading futures and were trading in a diversified manner.
We'll show you how you can construct and backtest a diversified trend following strategy without an extra $1 million dollars lying around to get started.
Most individual investors managing their own money don't trade futures. Why?
They require a lot of money.
Let's look at the most popular name in the commodities market - especially in light of current gas prices - crude oil. A single crude contract gives the owner control of 1,000 barrels of oil. As of writing, the price of oil is about $100/barrel, meaning you'd need $100k to own a single contract!
Now there are some caveats to this, namely leverage. The NYMEX doesn't require you to put all $100k up to control the contract, you can do so on margin. Depending on the oil expiration date, you may need to put up anywhere between 8-35% of the contract value to control it, and this can increase during times of high volatility.
If the contract you're interested in requires 30%, that's still $30k you need to control a single contract.
For trend following strategies, we like to be diversified for two reasons: protection and to catch those large outliers we love.
Protection comes from trend following's broad diversification. If you're just trading stocks, you can get hammered in a crash because the whole market tends to move together. It's not very likely that stocks and gold and bonds and the Swiss Franc and coffee and cocoa and...you get the picture...will move down together, making diversified trend following a safer bet in tumultuous times.
Further, we can improve our hunt for outliers by being broadly diversified because we're humble trend followers and we don't know where trends will emerge. It could be in oil, gas, cotton, coal, the Australian Dollar, uranium, gold, corn, Tesla, or anything else for that matter. If you need $30k to cover a single market, you can see how you're going to require a lot more money to cover all of these other markets so you can jump on trends when and where they emerge.
What are individual investors to do?
A Golden Era of ETFs
In recent years Exchange Traded Funds or ETFs have emerged to give retail traders access to a diverse array of markets around the globe. Today, there are ETFs for everything under the sun: commodities like the ones mentioned above, foreign markets such as Swiss or Singaporean equities, currencies, and more crypto ETFs coming on the market regularly. As with any strategy or fund, you need to understand the risks associated with any instrument you put into your portfolio (insert standard talk-to-your-financial-advisor-disclaimer here), something USO investors learned the hard way when oil went negative in 2020.
Moreover, ETFs have gotten cheaper as more competition has emerged. The expense ratio on an ETF is the key metric, which has plummeted from 2-3% (where many mutual funds remain) to as low as 0.03% per year. If that wasn't enough, you can get significant leverage out of ETFs for much less than it would to get that same leverage in your brokerage account.
Checking some retail brokerage rates, I found many will offer 3:1 leverage on liquid products like SPY (a popular S&P 500 ETF) and charge 8-10% per year. However, I could simply buy a 3X leveraged S&P 500 ETF like the SPXL with a 0.97% expense ratio. And, they've got a short version if you're interested in that too.
All this in a tax efficient package, ETFs provide a lot to love.
ETFs for Trend Following
You could construct a similar portfolio to the Turtles using ETFs, with some minor adjustments (good luck finding a Deutschmark ETF today) or build something from scratch. We're going to take the latter approach by accessing a number of leveraged, long and short ETFs and put them together into a trend following model. We'll show you how to construct this strategy step-by-step and run your own backtest!
Start off by going to the Strategy Builder to build and test your strategy without any code.
You'll be greeted by the settings tab where you can begin to select your strategy. We'll run this with $5,000 to see what kind of results we can get, but feel free to adjust the account size to fit your portfolio.
Start off by selecting a set of ETFs. We'll use the following pairs of long/short leveraged ETFs:
- SPXL/SPXS: 3x S&P 500
- TQQQ/SQQQ: 3x Nasdaq
- EUO/ULE: 3x Euro
- UGL/GLL: 3x Gold
- BOIL/KOLD: 3x Natural gas
We could use many more ETFs to get additional diversification, but this will do for now. Also note, that our free tier only allows access to the components of the S&P 500, so if you want to access the full universe of US stocks and ETFs, you'll need to have an upgraded account. You can find more about that here.
We'll start our backtest in 2011 because that's when the last of these ETFs came on the market, and still gets us over a decade of data to look at. This is one of the drawbacks of ETFs - we don't have as much data as we do for something like gold for backtesting, but we do what we can.
Next we need to set our Position Sizing and Position Management settings to control our risk. For this, we'll use a classic, ATR position sizing. I like to use a longer lookback period so I'll set this to 1 year or 252 trading days and leave the other default settings. For Position Management, I usually don't do much rebalancing after entering a position preferring to hold my position until I hit my exit, so I'll select No Risk Management for this option.
Entry and Exit Signals
I've had a lot of good success with a fairly slow, simple moving average cross over. For this setting, I'll use a 200-day and 250-day moving average to identify entry points.
To exit a position, stop losses or signal reversals work well (e.g. 200-day SMA < 250-day SMA). You could use these or a combination of both. For this model, I'm going to use a 200-day breakout to a new low, which we can set as shown below:
Feel free to try a stop (it works well for this model too) or a combination of breakout and stop or whatever else you like!
Once these are set, we can hit "Run Backtest" and take a look at the results.
As of writing, it will take about 20-30 seconds to fetch the data and run the test: about as much time as it took to build the whole strategy from scratch!
The strategy gets off to a hot start running up to over 4x returns in the first 4 years of trading. A lot of this is given back when the S&P 500 and Nasdaq take a dive in a tumultuous late 2015. The market dropped in August, rebounded, then dropped again into February 2016 before resuming it's upward trend. This kind of action can knock a long-term trend following system around - particularly if it is leveraged like ours.
Regardless our system recovered and managed to avoid a number of big drawdowns such as the COVID crash in March 2020, although because of its position sizing constraints, it didn't fully participate on the upside.
Also note, that the system was never fully invested, there's always some cash on the side. This could be deployed had we added some more instruments to trade or increased our target risk within our position sizing settings.
Looking at the backtest stats, we see that this turned in a healthy 13% annualized return.
There's a lot of good info in that backtest stats table. For one, we see that the system had over 300% returns with some major outlier returns with the biggest raking in a 539% profit! That's the kind of outlier you're hoping to hit with this kind of system!
The win rate is a bit low, even for a trend following system. If you look into the trades, this is partially driven by the fact that we have two different signals for entries and exits. It happens a few times where an exit signal gets triggered, so the system exits but the entry signal is still valid. In these cases, the system sells, then buys back in during the next trading day, often just to exit again shortly thereafter. A continuous system that uses the same entry and exit signals (just reverse them) would avoid this, or if we were to add another signal that requires us to wait say, 10 days, before re-entering a trade after an exit is triggered.
In all, there are lots of ways we could potentially improve this system, but the overall results are hard to argue with!
If you wanted to trade this, you'd just save the strategy and click over to your Bot Garage then turn it on (subscribers only) to get trade alerts whenever your system hits a trading signal!
Hopefully you see how quick and easy it is to get started with trend following with our no-code trading platform. Feel free to reach out or join us in Slack if you have any questions!