I’ve been taking a look at gaps recently, specifically putting some numbers on David Varadi’s Normalised Gap/ Lap Indicator.
First, it’s good to set some benchmarks with regard to the performance of gap fades on SPY (S&P 500 ETF). I’ve not looked at laps (where the market opens away from the close but within the previous high/ low.
Simple gap fade strategy:
- Gap up = open above previous high. Target previous high, exit at close otherwise.
- Gap down = open below previous low. Target previous low, exit at close otherwise.
Fade all gap ups on SPY since 2000:
Fade all gap downs on SPY since 2000:
David Varadi’s Gap/ Lap Indicator:
No two gaps are the same, which is where Varadi’s Gap/ Lap indicator comes in. This indicator normalises gap sizes so you can properly assess the size of a gap or a lap and therefore its impact.
Here’s the gist:
For brevity’s sake, here’s a summary of the results from the extreme values.
Summary for SPY Normalised Gaps:
- High probability of SPY closing gaps of any type, especially small gaps.
- Normalised Gap up:
- Best fades come with mid to high sized gap ups.
- Top quintile gaps run risk of being breakaways, though these too have performed well in recent years.
- Normalised Gap down:
- Best fades come with mid to low sized gap downs and top quintile gaps.
- Notes & further study
- Impact of overall trend?
- Relationship with DVB overbought/ oversold conditions?
- Add profitability by fading the right gap irrespective of other end of day signals.
- Caution on conclusions with low sample size and no frictions.
Next up…
I’ve the privilege of being involved with the next phase of the DV Plugin for XL Tool which has a nifty gap/ lap tool. I’ll report back on this very soon. I understand that ETF Rewind also contains a simplified version of the tool limited to select ETFs and without normalisation. (*)
Related posts:
