Geometric Balancing, Unlevered

Calculations are my own. Results are theoretical, do not represent returns that any investor actually attained, and are not an indicator of future results. Returns do not reflect trading fees. Dividends are re-invested.

The following begins a detailed look into Geometric Balancing. We start with the results of the unlevered portfolio. This portfolio is tuned to target a max 15% drawdown which it does (I will explain in another post how that works. ) Click on any chart to expand it for better clarity.

As discussed in the welcome post the strategy:

  • Is long only.  No shorting. 
  • Invests in 4 items:  the S&P500, the 30 year US Treasury Bond, Gold, and Cash.
  • Re-balances frequently. 
  • This strategy does not use leverage.
  • Applies cash to the portfolio via the Kelly criterion.
  • Uses a “factor of safety” in portfolio construction.
  • Does NOT use momentum in any way.

The performance of the strategy:The worst year is 2001 at -3.6%. The Best year was 1995 at 38%. Only 5 years end up losing money. The strategy often slightly lags the S&P 500 in the up years, and greatly outperforms in the down years. Even so, it beat the S&P 500 in 21 of the 41 years.

The following chart shows the geometric balancing vs the S&P 500 and a 60/40 portfolio (re-balanced yearly). You can see the portfolio beats the S&P, while never suffering painful, large drawdowns. Furthermore, geometric balancing handily beats the 60/40 portfolio while still producing much smaller drawdowns. It’s not just the best of both worlds, it’s better than both worlds.

Theoretical Performance Vs SP 500 and 60/40 Portfolio.

Calculations are my own. Results are theoretical, do not represent returns that any investor actually attained, and are not an indicator of future results. Returns do not reflect trading fees. Dividends are re-invested.

Next is a chart of the drawdown of the portfolio. The worst drawdown happened on September 21st, 2001. Outside of the post 9/11 tragedy, the worst drawdown was 12% in 1984 and 2009. Amazingly the strategy only crossed the 10% drawdown level 6 times in the past 41 years.

Drawdown

Calculations are my own. Results are theoretical, do not represent returns that any investor actually attained, and are not an indicator of future results. Returns do not reflect trading fees. Dividends are re-invested.

Below is the composition of the fund over the last 41 years. A few interesting items:

  • The portfolio will go 100% cash, the last time being early 2001
  • The portfolio has gone 100% stocks a few times, thought it’s not common. It can happen when bonds and gold are both very volatile.
  • The strategy only went full bonds once very briefly, in 1988.
  • Gold has never been above 32%. The highest it has ever been was in 2018.
  • The historical average is about 50% stocks, 35% bonds, 5% gold, and 10% cash. Although as you can tell actually residing at this average is rare.
  • It’s uncanny how the portfolio trades out of stocks leading into major corrections.
Portfolio Composition

Calculations are my own. Results are theoretical, do not represent returns that any investor actually attained, and are not an indicator of future results. Returns do not reflect trading fees. Dividends are re-invested.

Next is what I call the risked drawdown. It is supposed to crudely represent the likely maximum drawdown on the portfolio if I repeated the current portfolio for 40 years. As stated earlier, the strategy tries to limit drawdowns to 15%, therefore chart never falls below 85%.

Risked Drawdown

Calculations are my own. Results are theoretical, do not represent returns that any investor actually attained, and are not an indicator of future results. Returns do not reflect trading fees. Dividends are re-invested.

The portfolio aims for preferred return versus risk. Effectively it’s aiming for a high sharp ratio. At times the strategy will go into “safe” positions trying to meet this optimization. When it does, the risked drawdown rises well above 85%. It has generally been in this “safe” phase for the last 10 years, as seen above. The overall strategy works best when allowing some leverage due to this tendency. Leverage takes the safe portfolios and stretches them out to higher returns when appropriate. However some people are allergic to leverage, and anyway I find it best to evaluate a strategy free of leverage to start. But if you do apply leverage, this strategy produces even lower drawdowns and higher return than posted above.

7 Replies on “Geometric Balancing, Unlevered

  1. Really enjoying the blog.

    I’m curious to learn the ‘rules’ your system follows when determining the allocation to the various items (SPY, TLT, GLD, Cash). Can you please elaborate on this?

  2. There seems to be a 4 year stretch (roughly 1999 to 2003) where the portfolio returned nothing. At least my best guess from viewing that chart. Is that correct? I’m assuming it is based on the return chart you posted.

    The Harry Browne Permanent portfolio returned 5.4% (2.8% after inflation) CAGR.

    Basically that portfolio with a lot of trading to get higher returns but also a tad higher risk (although low compared to the extra 3% in returns.

    The weekly rebalancing (if that is what it is) seems impractical for most people.

    Anyhow, thanks.

    1. Yes you found the weakest stretch from 1999 through end of 2002. The portfolio lost almost a percent over that time. I have the permanent portfolio making %10% total over that time. So I will agree, over this 4 year period (less than 10% of the entire period shown) the permanent portfolio outperformed by a few percent a year. The rest of the time, Geometric Balancing outperformed the permanent portfolio by more.

  3. I have landed on your blog from Nick Maggiulli’s one.
    I’ve just started reading and it seems very interesting.

    I would have also added a simple table that splits the very long period into 5 periods of 10 yeras each (71-81, 81-91, …) and shows CAGR, Sharpe/Sortino, MaxDD for each period for your strategy towards the S&P500, 60/40, Harry Browne.
    So far the only weak period seems to be 1999-2000/2001, for some investors 1/2 years huge underperformance towards the market is enough to give up a strategy. 🙂

    Thanks for sharing all this math/statistic applied to portfolio investing. I’ll keep reading your blog to the end, I have got the feeling I’ll learn something cool.

    1. That’s a good idea about the 5 year period data. The 1999-2001 stretch is the worst, but it does better through that time than many strategies. 1999 does bother me though.

Leave a Reply

Your email address will not be published.