Yet realistically we can improve upon this model by looking at the long term upward drift expectations and rebalancing period. In real life fees may become an issue, but only if volatility is not great enough. Depending upon account size and size of fees, you may want to occasionally not rebalance unless the move is great enough, and you may also want to adjust the rebalancing period.
For example, on a typical day the market on average goes up about 52% of the time after adjusting it so all up days are equal to down days. In other words, your adjusted "edge" during "average" conditions historically has been 52% and as a result, you can position such that you have 52% in stock, and 48% in cash and rebalance daily. However, if your "rebalancing period" was more frequent, you should position much closer to 50%, and if it was less frequent, you should rebalance more often. But how much exactly?
Typical market
Rebalance
Daily 52%
Weekly 54%
Monthly 56%
Quarterly 58%
Yearly 62%
3years 68%
5years 72%
November1st - April30th period
Daily 54%
Weekly 57%
Monthly 61%
Quarterly 66%
--------
Since the period lasts 6 months there is no "yearly" re-balancing period. Yet, for comparison purposes only I ran the calculations anyways.
Yearly 74%
3years 83%
5years 88%
The May1-Oct31 period is close enough to a 50% proposition that I didn't bother calculating it.
Please note, I assumed a 5 day trading week and a 4 week trading month when making calculations and rounded to nearest percent. I also believe there was an improper calculation when using a 4 day period since I took all events that were up PLUS only HALF that were "flat" as I wasn't sure how to calculate this. Also, these number potentially are slightly generous due to every 1% loss requiring 1.01% gain to get back to even over a long period of time that may matter, however I was conservative in estimating 4 week trading month rather than 4.28.
A table if you prefer your data in this manner:
| Table | daily | weekly | monthly | quarterly | yearly | 3yr | 5yr |
| November1-April30 | 50% | 50% | 50% | 50% | 50% | 50% | 50% |
| typical market | 52% | 54% | 56% | 58% | 62% | 68% | 72% |
| May1-October31 | 54% | 57% | 61% | 66% | 74% | 83% | 88% |
| Strong Market | 56% | 61% | 66% | 74% | 83% | 92% | 96% |
| Very Strong Market | 58% | 65% | 71% | 80% | 90% | 97% | 99% |
| Super Strong Market | 60% | 68% | 76% | 86% | 94% | 99% | 100% |
I draw the line at 50% because the ideal way to play less than 50% probability of an upward move is either all cash, or even short plus cash which is much more complicated due to margin costs, and risks of ruin.
Commodities:
The probability of commodities outperforming stocks changes seasonally as well, but I don't have data available for far enough back to determine this easily. The data I found only goes back to 1992 and 11 data points per month, or even broken up into a 6 month period providing 66 data points is not all that statistically significant. Even if it were, there is a vast difference in correlation between base metals, rare earth and precious metals, agriculture, and energy commodities where in stocks the correlation between broad sectors is actually still very high. As such, the primary role of commodities for now is just providing reduced correlation to the portfolio, and individual commodities should be assessed according to expectations based upon success rate of indicators in specific conditions.
Really that's how all things should be evaluated and weighted according to confidence with excess income and cash for rebalancing, opportunity costs, stability and reduced volatility.... However the above table helps put a number on it and illustrates the importance of knowing your re-balancing period.
No comments:
Post a Comment