Improve Your Entry Price – Kaufman Signals

As a systematic trader I have spent a lot of time discovering ways to get a better entry price, all of them based on clear rules. A lot depends on your trade holding period, that is, whether you’re a short-term trader in for only 1 to 3 days, or a long-term trend follower with positions lasting 2 to 3 months.

Early on, markets had much clearer trend. Eurodollars, short sterling, Euribor, and other short-term rates have the strongest trends of all futures markets. Because they reflect Central Bank policy closely, they would move steadily in one direction, with very little noise. Trying to finesse an entry by waiting for a better price was a lost cause. It just kept moving away from you. Even with noisier markets, as the S&P, you would be able to beat the system entry price 3 of 4 times, but they were for small amounts. On the fourth time the market ran away and you lost all of the profit and then some.

In today’s markets, it’s easier to improve fills for long-term trend following than for fast trading. It’s easier to improve fills for index markets than for individual stocks. For improving fills, noise is your friend, and equity index markets have far more noise than individual stocks. There are stocks that can be significantly improved by waiting for a better entry point, but not those that are running away to the upside (or downside), such as Apple.

DEFINING THE RULES

To decide if we can improve an entry price, we need to have a system. Because a long-term trend will be a better example than a short-term strategy, we’ll use the following:

An 80-day moving average, representing a macro-trend system.
Buy signals occur when the average turns up; short sales when the average turns down.
Entries occur on close of the same day as the trend change, subject to conditions.
Exits are not subject to conditions, they always occur on the close of the same day as the trend change.

We’ll look at a number of tactics for improving the entry price:

The benchmark is to always enter on the close of the day that the trend changes.
Alternatively, on the day that the trend changes, close out the current position and place one of these orders:
1. Buy tomorrow at any price lower than today’s close (a “minimum” pullback).
2. Buy if tomorrow’s low is less than the prior n-day low (n = 1,2, or 3).
3. Buy if tomorrow’s low is less than the system entry price less a factor x 20-day ATR, where ATR is the average true range and the factor is 10%, 20%, or 50%. (The new entry varies with price volatility.)
Short sale signals are the opposite of the buy signals.

There are three simple but unique conditions resulting in seven combinations. We applied the trend strategy and the entry options to a number of individual stocks, equity index markets, and futures using 15 years of data.

SUMMARY OF RESULTS

Using SPY as an example, the first line shows that profits are all in the long positions. The Profit Factor is the gross profits divided by the gross losses and give you a reasonable reward/risk ratio. A value above 2.0 is very good.

When we enter on a “minimum” pullback the next day, total profits and ratio increase significantly, although the improvement is all because the short trades have a smaller loss. The three pullback tests (PB) show even more improvement, increasing the profits in longs and reducing the loss in shorts. The percentage of the average true range, which is a volatility measure, improves the long trades and hurts the shorts. We can nearly double the returns of SPY long-short trend trading by waiting for a better entry price (the 3-day pullback). We can increase the returns of the long trades returns by 20%, which is a better choice because all the shorts are losses and we will favor trading the long side. In terms of return to risk, the 50% ATR is best.

For individual stocks we’ve used Apple and J C Penny as examples. Apple seems to be always going up while Penny has been going down. This is a case where, if you wait to buy Apple, it can run away from you. Not so with JCP, where as much profit has been made on the short sales as on the longs. In both tables, the green highlighting shows those tests that were better than the benchmark. For Apple that was only the minimum pullback and the 20% ATR. For JCP it was everything.

The best candidates for entry improvement are markets with high noise. Noise is the frequent backing and filling that allows you to find a better entry price. For that, equity index markets are going to be better than individual stocks.

EQUITY INDEX MARKETS

We also found that different entry options could be very good on one index but not on all. In some cases it improved returns by 100%. But we want robustness in our selection, so a rule needs to work on a large number of markets to be acceptable. The rule that was most consistent was simply the minimum pullback. A summary of the index markets that we looked at shows that everyone had an improvement in the “All” profit factor, and in the absolute profits, showing that it improved returns more than it increased risk.

To see the results better, we’ve shown the percentage improve in the ratios for all the equity index markets as a percentage of the benchmark case. The improvement in the long trades varies from one case of -3% to +20%, but for the short trades it was consistently strong. We can explain that due to the difference between the long and short trade profiles.

Long trades are held much longer than short trades. Longs develop slowly. On the other hand, shorts tend to be highly volatile and stop quickly. It is difficult for a trend system to profit from a short trade because it is late getting in and late getting out, giving up too much of the potential gain. Of course, there are exceptions. By waiting for an upwards pullback after a short sale entry signal, we enter at a much better price. In some cases the trend might reverse quickly and no trade was entered. That would be even better.

The robustness of the entry rule that waits for a minimum pullback is seen in the consistency improvement in both long and short trades. But remember, this is for entries only, not exits. Finessing an exit can get you into trouble.