Monday, September 07, 2009

Further Conceptual Foundation for a Transitional Swing Trading System

My recent post illustrated how the great majority of days see the S&P 500 Index (SPY) trade above its previous day's high or below its low. Nearly half of all days, however, see the market close within its prior day's range. That suggests that false breakouts are common: many times, those moves to probe new levels of value ultimately fail. If one can develop filters to identify the fail days from the sustained breakouts, they would be well on the way of building a worthy intraday trading system.

Might the same logic apply to weekly price data, providing us with a possible swing trading framework? I went back to June, 2001 in SPY and found the following:

* About 56% of all weeks traded above the prior week's high; about 32% of weeks actually closed above the previous week's high;

* About 48% of all weeks traded below the prior week's low; about 22% of weeks actually closed below the previous week's low;

* A whopping 92% of all weeks trade either above their prior week's high or below their previous week's low; about 47% of all weeks ultimately close back within the prior week's range.

Again we see the common reversion: the vast majority of weeks trade outside the prior week's range, but nearly half close within that range.

If we combine these observations with the transition swing trade system outlined earlier, a very promising "mean reversion" swing trading framework starts to emerge--particularly if one uses the intraday transition patterns to aid execution.

That's all I have to say about that.
.

2 comments:

TelespallaBob said...

Wooow!
Doc, this sounds very interesting!
Look forward to see the final version.

As ususal... great job!
Many thank's

Bob

Thomas said...

Mean reversion applied to swing trading seems very promising based on the stats. It seems like most swing trading "advice" is based on getting long or short breakouts for a short term trend. The stats show that this will end up with net losers over time. Combined with a momentum system to buy/sell breakouts only when confirmed by other indicators this could have the makings of a nice overall swing trading system. Good luck with this and I will be anxiously following your work.