I'm not sure how to frame this out yet (perhaps others have ideas) but I am thinking of an accumulation indicator. The basic idea is this. Have you ever seen a market that went from "volatile" to almost a controlled, with a steady rise up. The qualitative thing you see is every single morning dip reverses very quickly. The second thing is that over a period of time there are no sustained pullbacks of any magnitude, an invisible hand guiding the market up. You can imagine how that kind of market feels for a short–every single short covering opportunity is thwarted prematurely.

Regardless, out of these conditions the qualitative hypothesis is that the price needs to accelerate before it can reverse or have a substantial correction. The question is, if defined quantitatively, might such an accumulation pattern show above average expected value. It is perhaps the flip side of the normal swing-type idea of buying a dip. Also, it might be helpful to only look at markets that have a positive drift.

John Bollinger comments: 

Fred Wynia's volume work addresses this concept quite well. The work
is proprietary and quite elegant/sophisticated, but the underlying
concept, that of measuring and comparing volume in swings, has been
around for a long, long time. As usual, the devil is in getting the
details right.

Gary Phillips writes: 

Ed, good luck trying to develop an indicator that is both robust and deterministic. Just a note however, if one only looks at markets that have a positive drift, back-testing results could be affected by said structural bias and rendered useless because they would only reflect the longer-term tendency of the market to go up.

Ed Stewart replies: 

Thanks. The idea as it stands is to complex to begin evaluating. I don't think I have captured the essential nature of the idea yet. I'm going to look if any specific elements of the idea on a stand alone basis. In terms of drift impacting results, that is very true. Drift needs to be incorporated in or it is pure futility. Many years ago when I was a random reader of the site I emailed in and Victor sent me a paper explaining a method that I still utilize, if I recall correctly. That ended up helping me tremendously. 

When I wrote the accumulation post, it was in large part based on watching the climb in IBKR over the prior few months and also similar observations on a short-time horizon. What do you know, IBKR has accelerated quite nicely. Up 5% today and almost 10% in prior 3 days. You can see the qualitative example if you look at a 3 month chart. No luck though, understanding the phenomenon on a systematic basis on the intermediate term. I've had luck with the idea on a shorter time horizon though. 

Gary Phillips adds: 

Most trend following systems have average win rates because of high draw-downs during whipsaw periods. The fundamental problem of most trend-following systems is that in order to deliver a high payoff ratio they must sacrifice a high win rate. If you try to increase the fraction of winning trades, the payoff ratio will suffer. So in effect, you would like to mitigate the negative effects of these problems by by combining a trend following strategy with a short-term trading system that would compensate for the negative trend following performance when markets are range-bound or mean-reverting. I am sure there are those that would argue that volume and volatility are both robust and deterministic indicators, but neither rising volume, nor falling vol, are necessary, nor sufficient, for the market to always trend higher, and even if they are randomly presented they do not necessarily establish timing. 





Speak your mind


Resources & Links