Live Proprietary Galton Mean-Variance Portfolio

The construction of optimized portfolios usually involves estimating optimization inputs from an historical sample of returns, repeated for every rebalancing frequency. The error between the estimate and what transpired, is disregarded at every iteration. The Galton-algorithm exploits predictability in these errors over time, to generate superior optimization inputs. Originally developed on monthly data for US stocks, I employ the method, in this setting, on weekly data for the Norwegian stock market. I find that this strategy produces portfolios that outperform not only other optimal portfolios but also naïve equal- and value-weighting schemes in the backtesting. I decided to take on the strategy as a proprietary, quantitatively driven portfolio. On this page you can track the portfolio weightings, gains, losses, as they happen in real time.

This particular version of the strategy rebalances every week (on mondays), but the performance is updated every weekday after market-closing (around 16:45).

2021 recap

Since its inception the 31st of May, 2021, the initial investment has grown by 14.04% compared to OSEBX's 8.57%.

Some metrics since I started measuring in week 29:

Risk: 13.54% vs 12.53% benchmark (annualised)
Return: 21.32% vs 13.52% benchmark (annualised mean return)
Reward/Risk: 1.527 vs 1.099 benchmark (annualised)
Biggest loss -4.54% vs -3.93% benchmark (instance)
Biggest gain 3.80% vs 3.33% benchmark (instance)
Dividend yield 2.82% (annualised, off season start)
95% VaR Breach 1 (VaR -4.32% vs loss -4.54%)
Beta 0.393
Alpha 15.9% (annualised)

Theoretical Weights

Below is a sheet containing the precise weighting scheme estimated by the algorithm. The realised portfolio weights are different due to liquidity, short sale costs, and transaction fees. I try to keep the general structure suggested here, without assuming unnecessary friction.