Parabolic Sar, how profitable is it?

The Parabolic Sar is an ‘all-in-one’ indicator, a stop-placement guide and a trading system which keeps the trader in the market at all times. The guiding principle behind the Parabolic Sar is that it solves a problem which has troubled traders for a long time.

This problem is that many existing trend indicators such as moving averages tend to give ‘late’ signals leading to missed profits during a new trend. The Parabolic Sar was developed as a direct attempt to overcome this ‘lag’ element.

The indicator appears as a series of dots underneath or above the price action, with a dot a certain distance under or over each bar. The dots start quite far away from the price at the at the beginning of a new trend and then move closer, forming a Parabolic-looking gradient. At certain points the dots cross over and move from under to over or vis versa, and this cross marks a reversal in trend.

The indicator was developed by Welles Wilder, the same analyst who developed the RSI.

The Parabolic Sar is a little different form other indicators in that it is intended as a complete trading system not just an indicator. When there is a cross-over of the dots and price action a trade is triggered in the direction of the new trend, so when dots move under price a long is triggered and when over, a short is triggered. When new trades are opened, all current open positions are also, obviously automatically liquidated.

The picture below shows Parabolic Sar in action on a chart of daily eurodollar with its characteristic sloping lines of dots.

As can be seen on the chart above the system seems to work quite well at a glance, getting the trader in and out of positions in quite a timely fashion, however, if we tested it rigorously, how well does it actually work？

Lets look at some trading results. Firstly lets take a daily chart of EUR/USD and back-test the Parabolic Sar over the last 12 years , using a standard lot:

On the daily chart Parabolic Sar is not profitable given the $-5887.10 loss, – which doesnt include trading costs -either. It is not terribly unprofitable either, as the results show that the strategy turned-over a gross profit of $283,288.30, and so the loss was marginal, however, it still failed to make any money.

Often strategies on daily charts are much more profitable on weekly charts because trends are more predictable, so lets look at the strategy back-tested using the same parameters on the weekly chart:

On a weekly chart it is much more profitable generating a substantial profit of $49,274.00 from a total of $143,843.70. The profit factor of 1.52 is also pretty good.

Now lets look at short time-frames – how well does the strategy work on intra-day charts？

First lets look at the strategys performance back-tested on 4-hour data.

The data from back-testing Parabolic Sar on the 4-hour chart shows the strategy is profitable , with a $49,626.70 profit versus $620,066.20 total profit.

This also indicates that the strategy does not necessarily deteriorate on smaller time-frames as it did going down from the weekly to the daily.

Despite being profitable on a 4-hour chart the strategy was not as profitable as it was on the weekly chart which showed an approx 49k return from 143k gross when the 4hr only showed a 49k return from 643k gross.

Finally now lets look at the hourly chart:

The strategy back-tested on an hourly chart it turns out is slightly more profitable compared to a 4-hour chart. The hourly gave a return of 121k from 1.193m, with a profit factor of 1.11 when on the 4-hr the profit factor was lower at only 1.09, however, trading costs would be higher using the shorter time-frame.

Enhancements

The strategy is successful on most time-frames, however, is there any way of making it better？

According the well-known technical analyst Martin Pring the best way to use the Parabolic Sar is by trading it in the direction of the longer-term trend only, therefore, if in an up-trend only the long signals are taken and if in a down-trend only the short signals are taken.

The next question is how to define the larger trend？ There are a variety of ways but let us try with a classic moving average, the simple 200-period MA. Lets look at the results for the different time-frames with the addition of the 200-period rule.

Below are the results of trading the Parabolic Sar over the last 12-years, only makes long trades, when price is above the 200-day MA, and not making short trades at all:

In most ways, the addition of the 200-day moving average improves the profitability of the strategy, which actually starts to make a profit rather than a loss on the daily chart, with a profit factor of 1.27.

Now lets test the same strategy on the 4-hour chart:

Lets look at the daily chart again but only trading short below the 200-day and not trading long at all:

Shorting makes substantially less money compared to using the indicator to go long. Now lets look at the same strategy on the 4-hour:

A little bit of an improvement but still not as profitable as going long.

Further Enhancements

Apart from the 200-day MA used as a filter, there are further enhancements we can test on the strategy. For example, one simple such optimization could be to only open long positions if the bar previous to the trigger bar confirmed the trend – ie green if opening a long trade and red if opening a short trade.

If the bar before the trigger bar was the opposite colour to the direction of the trend – ie red in the case of a long signal and green in the case then no trade would be taken.

The diagram below illustrates some examples of the enhancement so as to give a clear idea of how it works:

Now lets look at how successful the strategy is when back-tested as a trading model, firstly on a daily chart, back-tested on our 12-year EUR/USD data:

This enhancement appears to radically improve the performance of the strategy on the daily chart. Below are the results trading the strategy on a 4-hour chart:

Not bad either. Finally below are the results on a weekly chart:

The weekly chart shows a very healthy profit ratio of 1.78, which is the best of all the strategies. Therefore trading the Parabolic Sar with the addition of this enhancement appears to provide the best optimization for the strategy, with a weekly chart giving the best signals.

Conclusion

Overall we can make several observations about Parabolic Sar and its effectiveness if traded on EUR/USD. The first is that it seems to work best on the weekly chart: the un-optimized strategy gives a profit factor of 1.52, which is second best, and with the help of the second enhancement, in which trades are only taken if the previous bar confirms the trend, delivers the best results, with a profit factor of 1.78.

Secondly we can say that it is overall more profitable to trade the Parabolic Sar by going long rather than short. Even with the aid of a 200-day moving average to filter out counter-trend signals the shorting version of the strategy only made profit factors of 1.15 and 1.06 on the 4-hour and daily charts respectively. The un-enhanced Parabolic Sar results also showed consistently lower profits on the short side, with the 4-hr delivering 1.03, and the daily chart actually showing a loss which was the sole reason for the strategies unprofitability overall on the daily. Weekly showed the best results for shorting, with a profit factor of 1.30 for the un-optimized strategy.

Overall, therefore, if you are going to use the Parabolic system it is probably best to use it as a way of getting long on the weekly charts.