A.I.M. Users Bulletin Board

[ls7550]

Hi Tom

I've struggled with the idea of what weight each sector should have in the total portfolio

Initial equal weight will tend towards its own cap weighted over time. If you cap weight at the offset then in effect you're playing a momentum approach, investing more into the sector(s) that have yielded the current cap weighted levels. In the absence of being able to predict which may be the better performer going forward, best perhaps to just back the field evenly. If for instance you overweight tech because that is what had done well to date, then that's less invested in others, and it could be one of those others that is the better performer in 5, 10, whatever years time.

The characteristic is fractal. Look at say 10 assets over a intra-day and ranked worst to best will tend to see the single good case outcomes being tall/up enough such that the majority of individuals lag the broad average of the whole. That can be scaled up to 1 year holdings, more or fewer assets, and even for entire countries over decades. One or more will tend to do well, others poorly, and the majority lag the broad average because the right tail best case will be taller (perhaps +200%, +300% or whatever) than the left tail worst case (that has finite downside of -100% at most).

As for rebalancing back to equal weightings, well overall generally it makes no difference to rewards in general whether you do or don't. What rebalancing back to equal weightings can do is to potentially reduce volatility, so mathematically has the better Sharpe Ratio. Seems logical if you've held 10 assets say long term and the portfolio had transitioned to being 50% or more of the entire portfolio value concentrated into just two holdings to de-risk the over-concentration into those two stocks/holdings.

John Maynard Keynes (a British economist) accumulated over 100 pieces of artworks, long term (many decades) the total returns pretty much compared to had he invested in stocks with dividends reinvested, but where over half of the value was within just a couple of pieces, and over 80% of the total value was in just a handful. I guess the sensible choice would be to sell the most valuable, load up your cash/drawdown holdings with some of the proceeds, and buy a wide range of other lower priced pieces in the hope/expectation that one or more of those might be the 'big up' winner(s). But equally it doesn't seem to have bothered not doing so. Some say run-winners/cut-losers, others look to mean reversion, neither is consistently the correct choice, so a good choice is perhaps midway between the two. Only after the event will it become apparent which was the right choice.

Overall my guess would be to initial equal weight, rebalance periodically those that had risen to 2 times the median average back down to being 1.5 times the median value. Or something along those lines.

Regards. Clive.