Short term reversal - Jegadeesh (1989) Short term reversal 1934-1987

General
1

Short term reversal is said to be one of the better strategies - Jegadeesh (1989) Short term reversal STreversal 1934-1987.

I have tried various approaches to see if they provide better returns in my strategies.

Some of the approaches I have tested include:

TotalReturn4W
(close(0)/close(5))/pctdev(5,1)
(close(0) - LoopMin(“low(CTR)”,10,1)) / (LoopMax(“hi(CTR)”,10,1) - LoopMin(“low(CTR)”,10,1))
(Ln(Close(0)/Close(5)) - Ln(Close(0)/Close(251))*(1/51) ) / LoopStdDev(“Ln(Close(CTR)/Close(CTR+5))”,51,0,5,0)
rsi(4)
rsi(8)

However, none of these approaches seem to contribute significantly. Has the effect been arbitraged away? Or does anyone have suggestions for other rules that can be tested?

2

I’ve had some luck with:
loopmax(“close(ctr)/close(ctr+1)”,30)
(Ranking: lower & universe).

You could also try to extend it like this:
loopmax(“close(ctr)/close(ctr+1)”,30) - loopmin(“close(ctr)/close(ctr+1)”,30)
(Ranking: lower & universe).

3

I use close(0)/close(20) plus a sprinkling of close(0)/close(10) and close(0)/close(30) (Weights 5:1:1). I find it to be quite strong, removing it takes away more than 6 percentage points of alpha from my favourite ranking system.

1 Like
4

Thank you!! When trying to test it Im getting this error?:

image

5

Perhaps the problem is caused by the quotation marks? Try copying this formula instead:
loopmax("close(ctr)/close(ctr+1)",30) - loopmin("close(ctr)/close(ctr+1)",30)

6

I use a quirky version:

weeksintoq*(close_d(0)/close_d(20)-close_d(0,#sector)/close_d(20,#sector)), lower values better.

If you see a big price move coincident with an earnings announcement, ignore it because it is likely to continue. If you see a big price move that’s quite some time since the earnings announcement, it’s likely to reverse. You should always compare price reversal to industry or sector prices, because short-term momentum in industries and sectors is a powerful force. What you’re looking for is price movement that is not explainable by earnings announcements or industry/sector moves. That is the kind of movement that is most likely to reverse.

Also, you should be aware that using rules like these significantly increase the amount of trading you do and any benefits may be offset by the additional trading costs.