With this post I want to share a simple GAP fading strategy. In my never ending desire to find complementary strategies to my existing portfolio of strategies. I looked into GAP fading. What’s GAP fading: Assuming a stock or ETF opens higher than yesterday’s close, you got a GAP. To fade a GAP you essentially take a mean-reversion trade. So one is expecting that the market is likely to trade lower, hence trying to fill the GAP.

Why GAP fadding? My existing strategies tend to trade between 2-10 days. Typically GAP fading strategies are shorter. Furthermore I decided to look into fading an UP-GAP only, hence taking on a short trade. Because my existing strategies take on more longs than shorts.

About the strategy

David Varadi has posted an idea to normalize the GAP. The basic idea: different instruments show different inherit volatility. So to calculate an UP-GAP I subtraced yesterday’s close from today’s open. This got then divided by ATR(5). To compare the POWER of an up-gap it get’s PERCENTRANK’ed. This way you get an indicator that is able to capture an UP-GAP and compare it across very different instruments.

What to trade

Taking on a short trade after a signficant UP-GAP is a counter-trend trade. My assumption is, that ETF’s show a higher tendency to mean-revert that individual stocks. Hence I created a portfolio of about 33 ETFs made of Sector SPDRs and most liquid MSCI country ETFs.

Strategy – Results

  • Entry rule: enter at the open when the UP-GAP is within the TOP quantile
  • Exit: at the close
  • Positionranking: In case of multiple entries the two with the strongest UP-GAP will be picked.
  • 50% equity per position
  • No commission / no slippage

return.png?w=300&h=278

(click on the picture to enlarge it)

Review

The strategy performed EXTREMELY well until end of 2009. By the beginning of 2010 it literally stopped working. I’m asking myself: what caused the sudden death of the strategy? Can it be explained due to shrinking volatility post financial crisis? Partially, already in 2009 volatility was shrinking but the strategy performed very well (122% CAGR).

I would like to get your thoughts why the strategy stopped working. Please drop a comment or sent me an email hassler.frank (at) gmail.com.

– Frank

AmiBroker – AFL

SetBacktestMode( backtestRegularraw ); SetTradeDelays( 0, 0, 0, 0 ); SetOption( "CommissionAmount", 0.00 ); SetOption( "AllowSamebarExit", True ); SetOption( "MaxOpenPositions", 2 ); SetOption( "MaxOpenLong", 2 ); SetOption( "MaxOpenShort", 0 ); SetPositionSize( 50, spsPercentOfEquity ); RoundLotSize = 0; Short = Cover = Buy = Sell = False; ShortPrice = Open; CoverPrice = Close; ATR5 = Close * ( ATR( 5 ) / MA( Avg, 5 ) ) ; Gap = IIf( Open > Ref( Close, -1 ), ( Open - Ref( Close, -1 ) ) / Ref( ATR5, -1 ), 0 ); Rank = PercentRank( gap, 252 ); Short = Rank > 75; Cover = True; PositionScore = rank;

1267 1267 1267 1267 1267 b.gif?host=engineering-returns.com&blog=