Trading for a living requires steady returns. Equally important, and often overlooked, is that you must avoid big losses to stay in the game. Gains and losses are asymmetric and the math is well known – a 10% loss requires an 11% gain to break even; a 20% loss will need gains of 25% to fully recover. Looking at long-term market trends can help put the odds on your side for successful trading.
It’s easier to make money with long trades in a bull market and short trades are more likely to be profitable in a bear market. Simple filters can help traders take full advantage of a trend. A simple strategy for short-term traders, like increasing position sizes on trades that are in the direction of the trend, could lead to bigger gains. An even simpler approach would be to require that trade signals are confirmed by the filter before taking the trade.
Testing that idea should show whether or not it works. These tests will be completed using a diversified basket of futures contracts. It’s simply not possible for small traders to earn a living trading stocks, even if they are day trading with the maximum amount of leverage possible. Futures are highly risky but with a good plan the risk can be managed and the high degree of leverage will make it possible to earn a living from a small account. These tests will use a single contract in each of seven markets: cotton, copper, feeder cattle, sugar, crude oil, the US dollar index, and five-year Treasury notes. To trade all of the contracts, the exchange minimum margins would be about $27,000. Smaller traders could start with a few of the contracts and add more as their account grows. The dollar is a low margin contract that shows strong trends and is a good market to learn the basics of trading.
These tests will cover the time period from January 1, 2000 through August 29, 2011. This time period represents a fairly typical market with both upside and downside volatility in each of the individual contracts, and long periods of time when the markets didn’t move much at all. We’re testing with end of day data and using the next day’s opening price for trade execution. Small traders usually have day jobs and can’t monitor the markets continuously, so these tests recognize the limited time they can devote to trading. Trading costs of $45 per round-turn (a buy and sell transaction) are deducted from each position. This is much higher than the commission you’ll pay with your broker, but back testing should set a high hurdle for a system to help present realistic assumptions of what you can make in the markets.
A good strategy that I keep writing about is to buy new highs and sell new lows. In this case, we’ll enter long when the price closes at a new 20-day high and reverse to a short when price breaks to a new 20-day low.
It’s a simple strategy with average returns of about 25% a year. The maximum drawdown amounts to about a third of the total profits, which is the upper limit of risk in an acceptable trading system. On average, there is about one trade to place each week, so this idea could easily be used by part-time traders.
The same strategy can also be tested with a trend filter. Signals will only be taken if the 200-day moving average confirms the price breakout. Long trades will require a new high while the close is above the 200-day simple moving average, and short signals will only be taken when prices make a 20-day low while below the 200-day moving average.
Returns are a little less, with an annualized gain of 24.85%. But, the drawdown is dramatically reduced and is now near 12% of the total profits. Trading frequency is also reduced, with the system now averaging 20 trades a year instead of more than 50 that we had without the filter. The small decrease in profits comes with a very significant reduction in risk. To me, the tradeoff is worth it and I would use the filtered signals.
The biggest problem with this strategy is that only 25% of the trades are winners. We’ll try adding one more filter to see if we can improve the returns. Sometimes new highs signal the end of a trend rather than the beginning. If we only take signals that occur after a pullback, we should avoid the spike highs and lows that often mark reversals. This filter will be simple, rather than optimized and will only take signals if the close the previous day was opposite to the direction of the signal. That sounds confusing, so I’ll try to explain it better with some very light math:
– Only take long signals if yesterday’s close is less than the close two days ago
– Only take short signals if yesterday’s close is greater than the close two days ago.
This idea is being tested in its simplest form, but traders can work on different parameters to find what works best for them.
With the added condition, we get 33% winning trades, a small improvement. But we see significant improvements in other system testing numbers. Profits increase and risk, as measured by the maximum drawdown, decreases. Annualized returns are 27.21% a year, and max drawdown is only 11% of the system profits. There is only one trade a month, on average, so the system lets winners run and minimizes trading activity.
Simple ideas work in the markets and this trend following strategy could easily form the starting point for small traders to learn about the markets while growing their accounts.
By Michael J. Carr, CMT