Trading Exit Strategies - The Beauty Of Time-Based Exits
If you've read many of the countless books on trading or listened to some of the many market 'gurus' out there you'll no doubt have heard the theory that entries don't matter, only exits do. The idea goes that entries don't provide a tradable (or any) edge, and since picking exact tops and bottoms is nearly impossible, almost all trades will be in profit at sometime so success doesn't depend on when you get into a trade, it depends on when you get out of it. This is not a view that I agree with and the profitability of systems that use time-based exits are my response to these claims.
Broadly speaking, all the mechanical price-based exit strategies that I have come across fall into one of four categories, or are a combination of several of the four. These mechanical exit strategies are –
1. A Trend Reversal Signal,
Trend reversal signals are usually used as part of long-term trend following systems; the aim of these systems is to make the most of large trends by 'riding the trend until it ends'. This tends to produce good results when the trend lasts, but more often than not trading systems that use these exit strategies produce more losing trades than winners; with the size of the winning trades being at least several times larger than the losers. The reason why most trades using these exits are losers is that these exits enviably involve 'giving back' some of the profits made as they wait for the price to reverse and go against the trade for sometime before they produce an exit signal – most trends don't last long enough after they've been detected to overcome this, although the few that do are usually very big. Some examples of trend reversal signals might be a new 20-day low during a long trade or a new 20-day high during a short trade, or a moving average crossover against your position etc...
2. A Lack of Follow Through Signal,
Using a lack of follow through signal as an exit strategy tends to produce a higher winning trade to losing trade ratio than using a trend reversal signal, but it often fails to fully exploit large trends. A lack of follow through signal is exactly what it's name suggests – an indication that the trend is not following though (i.e. continuing to move in favour of your trade). An example of a lack of follow through might be the price failing to make a new high in 20 days during a long trade or the price failing to make a new low in 20 days during a short trade. This type of exit tends to produce more winning trades than a trend reversal signal as the price doesn't have to move as far against your position to produce an exit (an exit can be triggered by the market just moving sideways), but it does have a tendency to cause trades to exit before the trend is over as sometimes the market starts moving sideways for a great length of time before continuing to move in the direction of the long-term trend.
3. A Favourable Price Signal,
Exiting with a favourable price is the polar opposite of a trend reversal signal, as the name suggests, the exit simply involves taking a relatively favourable price when you have it. Such exits are the logical conclusion of 'entries don't matter, only exits do' thinking. If entries were random or could not produce an edge then the logical time to close a position would be when one had a favourable price or when the market had moved to far against your trade. An example of these exits might be to close a long trade after a significant high or close a long trade when the price closes higher for three consecutive days, and vice versa for short trades. Favourable price exits tend to produce far more winning trades than losers and almost never take full advantage of the big movements and long-term trends. These types of exits are typically used in systems that buy dips and sell into mini-rallies. They work best when the market is 'choppy' and tend to work best on markets like stock indices rather than markets like Forex and commodities that tend to trend well.
4. Time-Based Exits,
Time-Based exits are the simplest forms of exits and, in my opinion, one of the most under rated. If an entry did not have an edge then using time as an exit strategy could not produce a profit in the long term and a trader would slowly lose money as the costs of trading slowly consumed their entire trading account. It's only because entries do have an edge and do matter that time-based exits are profitable. A time-based exit is simply the decision to exit a trade at a fixed point in time in the future regardless of what the price does after the trade is entered (although these exits are usually used with, albeit it large, stop losses). The theory behind these exits is that the trend is your friend but it needs time to work and you'll never know when the ideal time to exit is. So when the market is going up you enter long, give it x amount of time to work, the get out, and don't complicate it with anything else. This remarkably simple exit tends to work very well over times of about 20 days or so. Systems that use these types of exits tend to produce slightly more winning trades than losses (usually trades tend to win about 55% - 60% of the time) and the winners tend to be slightly larger than losers too. The above 50% winning rate and the winners being slightly bigger than the losers comes purely from the entry's edge. If there were no edge in the system's entry, time-based exits would, on average, produce winning and losing trades in equal number with the winners being of equal size to the losers (unless a good stop loss is used).
Conclusions,
None of the four exit strategies are without merit. Trend reversals work very well in strongly trending markets, particularly in long-term trend following systems that aim to take advantage of large moves over several months or longer, and especially in Forex and commodity markets. Using a lack of follow through also works very well during these strong trends, this exit often fails to fully capitalise on them but it does produce a higher percentage of winning trades than a typical trend reversal exit as it generally involves 'giving back' a lot less of the profit. Favourable price exits work best in 'choppy' sideways moving markets, especially stock market indices, and they tend to produce a far higher percentage of winning trades than losing trades. But the most under valued is probably the time-based exit. A time-based exit's beauty is in its sheer simplicity; it's the simplest of all the exit strategies and often the most effective. Providing the system's entry provides a statistically significant edge then a time-based exit is more likely than not going to produce a winning trade and the winnings trades will also tend to be larger than the losers too. With a completely random entry that is no better or worse than a coin toss a time-based exit would (on average) produce a winning trade to losing trade ratio of 50/50 with the size on the winning trades and the losing trades being exactly equal (assuming no use of stop losses). However, with a statistically significant entry edge and time-based exit the win to loss ratio will climb above 50/50 (often as high as 60/40) with the average size of the winning trades to losing trades climbing by around the same amount. The idea behind a time-based exit is very simple, it goes something like this: 'the market is moving in a certain direction and is more likely than not going to continue moving in that direction for sometime, but we don't know for how long. So we get in, give it x amount of time to work then get out. And that is it' – so simple but often so effective, the beauty of time-based exits.
