A bunch of people got in touch about the recent 'Duality' strategy.

Frankly, I see a result and post it up here, thinking these ideas through as I write and as you read 'em.

I used current DJIA constituents to backtest the strategy, but mentioned the need to test with random portfolios. Randomness is very tricky, so instead, I thought I would test with all the stocks which are currently trading but dropped out of the DJIA at some point over the last 100 years or so.

Here they are:


These companies carry a lot of baggage.

If we equal weight each, and backtest since GM started trading again in 2010 we see a Sharpe of 0.145.

If we apply the Duality Strategy to the same stocks we see a Sharpe of -1.184!

Which is pretty much the opposite of what we see when we apply the strategy to the current constituents of the DJIA.

Crazy, but expected, Duality really just reduces volatility, a lot.

What's the take away?

If we are stock pickers, and have conviction in our methodology, the Duality Strategy accentuates the positive. But things get pretty grim if we hold onto losers.

When I remove GM from the mix we have over 10 years history to backtest with. Things are less bad.

Equal weighted Sharpe is 0.09. Duality is -0.14.

Surprise, surprise! Duality is no silver bullet against shooting yourself with some of the worst stocks of all time.

Less extreme portfolios need to be tested.

My hunch is, as long as your stock picking portfolio beats your chosen benchmark on average, you stand to see a decent improvement with Duality.

More to come.

[Thanks to Hong-yiu Lin for listing out a bunch of terrible stocks as well as giving me a gentle chiding(!) which spurred me to write this]