Follow Up to Qusma’s IBS and Relative Value Mean-Reversion
Qusma is one of my favorite new blogs. This is a follow up to their post IBS and Relative Value Mean-Reversion in which they test a pairs trading strategy based on various (mostly international) ETFs under-reacting to changes in the US market.
Read Qusma’s post as a primer. I’ll contribute to the discussion by adding an equity curve from 1997. First, excluding trading frictions…
[logarithmically-scaled, growth of $1, frictionless]
Strategy rules (using Qusma’s universe of 32 mostly international ETFs):
If SPY IBS (Internal Bar Strength) < 50% and ETF IBS > 80% at today’s close, go long SPY and short the ETF at today’s close. If SPY IBS > 50% and ETF IBS < 20%, go long the ETF and short SPY. Close all positions at the following close.
Allocate 25% of the portfolio to each pair unless more than 4 pairs are active. If more than 4 pairs are active, split the portfolio evenly amongst the pairs.
One last bit of slickness – rather than using Qusma’s equal dollar weighting, weight the pairs to equalize the standard deviation of daily returns over the last 63-days (1-quarter). That means the ETF will usually receive a lower weighting than SPY (because the international ETFs tend to be more volatile).
Of course, transaction costs and slippage would be a major concern with such an active strategy, so here’s the same test with 0.10% round-trip cost included.
[logarithmically-scaled, growth of $1]
As Qusma notes, the strategy is just a historical curiosity at this point as all of its mojo has clearly been traded out.
Seeing a strategy with the type of equity curve this one exhibited until 2003 always gets the nerd juices flowing. But I’ve sliced and diced the strategy a dozen ways to better understand if some change would make the strategy productive again (like trading the pair when the ETF over-reacts to the SPY rather than under-reacts) to no avail.
Quite a bit of work went into putting the strategy through its paces, so I thought I would share to help get other readers’ nerd juices flowing.
. . . . .
Filed under: Trading Strategies | 2 Comments