2019 Year in Review and Portfolio on 1.3.20

We’ve closed another year, did Geometric Balancing perform like a fancy expensive hedge fund? Did it produce an exceptional return, with little volatility? Let’s find out.

Actual Geometric Balancing Results

The results below show actual returns from my account. This is the same portfolio I update and publish each week¹, and the same strategy that runs live at the top of the blog. The following chart shows the actual results for Geometric Balancing without leverage from the beginning of April through the end of the year. Before April I traded with leverage. Blending the results wouldn’t properly show the strategy’s performance. The results are strong without leverage, and I don’t want to confuse the analysis.

You can see how Geometric Balancing slightly trails the S&P 500 but outperforms the total stock market (VT) and international markets (EFA). It’s also easy to see the smoothness of the returns, much less choppy and volatile than the others.

Interactive Brokers produces a risk analysis which I’ve included below. The analysis covers daily returns from early April through the end of the year. The realized metrics for Geo. Balancing match closely to those predicted in March: Nearly the same return as the S&P 500, with half the standard deviation and far less drawdown (the VAMI pretends all funds start the period at $1,000, and therefore shows the % return).

Simulated 2019 Returns

To show estimated full year results, I took the the model’s prediction for January-March and added it to the actual investment results from the April-December.²

The combination comes to a 21% annual return, a Sharpe ratio of 3.2, and a standard deviation less than half the market. Every statistic easily surpasses the S&P 500, except for total return. All without any leverage.

Great, But You Still Lagged The Market by 10%

You may say, ok this is a great Sharpe ratio, clearly superior to the S&P 500. But Geo. Balancing still lagged the market by over 10%, which is a lot.

Yes, that is true. However the the Jan 1-Dec 31 range is extremely favorable for the S&P 500. It’s a bit cherry picked. The market in 2019 moved almost perfectly from an intermediate low to all time highs. Most strategies under perform when compared to one moving perfectly from a low to a high. Don’t be confused by the particular time window.

From the above chart, its clear the out performance for the S&P 500 comes almost entirely in the first 3 months. Due to continued high volatility, Geo. Balancing often stays conservative when exiting steep market drawdowns.

In December 2018 the S&P 500 lost over 9%. Geo. Balancing gained nearly 1% during the mini-panic, a 10% out performance. The strategy lagged so much for the first quarter because it spent much of that time protecting against the possibility of a resumption of the drawdown. Holistically the the first quarter of 2019 is just the second half of a cycle.

Using a 13 month window, the returns of Geometric Balancing converge to equal those of the S&P 500, with half the volatility, and a third of the drawdown. Exactly as expected.

Lets Talk About The Lag Out of Bottoms

I would love to get rid of the lag coming out of bottoms. I’ve studied them trying to find a way to eliminate it. I’ve only found two methods:

• Increase the risk tolerance. This is fine, but the risk tolerance is set where it is for a reason. If this is increased, then the drawdowns grow with it, a typical trade off for all strategies.
• Improve the model inputs. This should work too, but is easier said than done. This is where I’m focusing most of my efforts for improvement today.

Here’s the thing about the lag: Even though Geometric Balancing lagged the market by 10% for the first quarter, the strategy still gained 7%. Many, strategies capable of reducing downside losses would have left the market entirely during early 2019 and experienced none of the rally.³ I prefer a strategy which still experiences as much of the upside as possible.

Did Geometric Balancing Live Up to Expectations?

Yes it did.

And interestingly, the hedge funds I like to compare the strategy to did not have good years. Geometric Balancing: just 4 assets to achieve hedge fund like returns? Maybe it should be just 4 assets and better than hedge fund returns?

Either way, It should be clear, you don’t need the complexity, high fees, and limited access of fancy investment funds to produce stellar returns. Just combine the wisdom of the Kelly criterion with the power of rebalancing. By this time next year, maybe you too will report similarly spectacular results in your portfolio.

January 2nd, 2019 Portfolio Rebalance

To close out, here is the update for the week.

Up nearly 0.6% for the week. It’s been a strange few weeks for the strategy. It keeps on moving early in the week into stocks, only to return to nearly where it started by Friday. On Jan 3rd, 2020, the portfolio rebalanced to :

68% SPY , 16% TLT , 16% GLD

1 – I try to make these trades at the end of the day. Obviously, I can’t make the trades exactly at the end of the day on Friday, but I do try.

2- The model’s results from April through December closely matched the actual results. The model was 3/4% higher before any fees and traction costs. After transaction costs and ETF fees, the realized results lagged the model by about a 0.33%. As mentioned above, it’s never going to match perfectly due to the fact I can’t trade at the actual close. The variance between the model each month was not consistent. Some months my strategy won, others, the model won. Therefore for now, I’m chalking this up to implementation randomness.

3- As an example, an all or nothing trend following strategy on the S&P500 with a 200 day moving average would have bailed out in early December 2018 after small losses, and not re-entered the market until mid February. Geometric Balancing gained 5% over this time.