Posted on

Ultimate Black Swans

Hilarious: I thought 1929 was the ultimate absurdity until I went to the bank of England website and took down their stock market data (monthly) from 1709.

I cleaned up the dates (dates before 1900 can be quite a laugh) and ran a simple trend following system against the data. A bog standard asset allocation system such as you see on many websites like  Gary Antonacci’s Dual Momentum.

And lo and behold the great South Sea Bubble of 1721 hit me in the face.

Dear old Gary has done a good job back testing his Dual Momentum system back to 1971 with MSCI Index monthly data.

But he really should take a look at the South Sea Bubble.

7 thoughts on “Ultimate Black Swans

  1. I think this paper more fully evaluates the potential for the DM system. – diversifying across asset classes.

    1. Too little history from Gary. Sure, I believe on balance it works. But not all the time and certainly not during a Black Swan when you either get lucky and are on the right side of the trade or you don’t and you are screwed.

      Running bactests on 300 years of data show you can and will get screwed. Eventually. Remember Weimar Germany, Imperial Russia and early 20th century Argentina.

      Could the dinosaurs have diversified and avoided extinction? Probably not.

      Can you diversify and avoid financial extinction?

      You have to try. The mistake people like Gary make is in putting too much faith in a single system over a very limited timeframe. If I recall correctly he used data back to 1971.

      Not sufficient to alert you to the pitfalls. For instance a binary system is not wise if you want to avoid large drawdown.

      I could drone on and on but I won’t. No problem with GA at all. I just believe he is overconfident.

    2. Gold, silver , wheat, timber. None of the correlations or lack thereof are dependable over time.

      All is correlated in a crisis these days. Or you get caught out such aa the recent flight to the yen not CHF over Brexit.

      My point? You can’t afford 100% exposure to risk assets at any time. Don’t be binary: it’s not bonds OR stocks. It’s bonds and treasuries and a portion of stocks when the going is good and a retreat to bonds and stocks only when the going gets tough.

      And NOT restricted to issuance by one Sovereign! Even the Americans should realise empires eventually fall.

    3. It goes back to basic physics: even if we live in a deterministic universe, given our current state of knowledge we are unable to forecast the future.

      No one has THE answer. No one has discovered the philosopher’s Stone. Certainly not I.

      Retain your scepticism and deep waryness at all times. Don’t pay too much attention to backtests – they usually turn out to be more or less useless.

      Almost everyone is trying to sell something and their advice will be clouded by that aim.

      Wish I had something to sell but I don’t think I so.😂

    4. By the way – I have not published a back test on this. I have done many at home and many look promising even just using stocks, a ten year bond and treasuries. In varying proportions.

      But what both my tests and my common sense and my experience all tell me is that is won’t all turn out as rosy as you think.

      I have seen and read so many of these type of papers and they now give me quite some amusement. They are not “wrong”, the are not badly researched, or misleading.

      They are just hopelessly optimistic.

      In my view. Obs.

      1. I generally agree with all you comments and I know from tradingblox days that you have a lot of backtesting experience. No argument there.

        I do think that using both relative and absolute momentum is a very promising approach. The biggest advantage being the much reduced drawdowns an thus being able to stick with the system. That one attribute, reduced drawdowns, makes the approach worth further investigation.

      2. Agreed. I will publish my test on this data and on US data going back to 1870. A binary approach (all stocks or all bonds) produces VERY scary drawdown in 1721 and 1929. An approach which never invests more than 50% on stocks produces very acceptable results. Totally agree with your point on following the system. Totally agree.

Leave a Reply