Bubblesofearth wrote:OhNoNotimAgain wrote:
Terry, you know the maths of weighted averages better than the next man. The bulk of the market return comes from a couple of dozen stocks. Get those right, with appropriate weightings and you are sorted.
Totally agree. I've looked back over a few time periods and, after a careful review of share performance tables, have managed to do exactly what you say.
All I need to do now is the same but without the performance tables, i.e. for a portfolio launched today rather than a decade ago.
BoE
The appropriate weightings is to initially equal weight holdings, unless you opine you can pick those that will be the best/right-tail holdings in advance (as simple as that may seem, more likely you can't as the best forward time cases more often are surprising). After that it doesn't matter if you do or don't rebalance, other than non-rebalanced tending towards having more concentration risk in having found its own cap weighting and where you likely end up at a point where a initial portfolio of say 20 stocks initially weighted 5% each having become 80% of the portfolio value held in a handful of the stocks (the best performing stocks/sectors across that period of time). Look at rebalancing as a possible case of selling/reducing out of your best performers to add to laggards that may continue to lag.
Buying initially into the ongoing market cap weighting is weighting more initially into what sectors/stocks had already been the historic best/right-tail sectors/stocks, where there is no guarantee that that momentum will be sustained into the future. Leaving less weighting having been bought of alternative sectors/stocks that may very well be the best/right-tail in forward time.
When I last looked at HYP1 (non rebalanced initial equal weighted), compared to TJH HYP (rebalanced/equal weighted) 2000 to 2020 inclusive, total returns (accumulation) the linest(ln ... (slope of exponential trend line) values were more or less the same, as were the motions of the total return lines. The standard deviation in yearly total returns was higher for TJH-HYP than for HYP1, so for those particular cases the risk/reward was better for HYP than for TJH-HYP (same/similar gain with less volatility (where volatility is considered as being a 'risk')).
Another real world example is US LEXCX that started with 30 shares in the 1930's, bought and held (no rebalancing), that broadly has compared to the market index (rebalanced).
You can do OK/well either way, rebalance or not, however a set of 20 or 30 stocks sampling based approach will have some pick above average holdings, other pick below average holdings, deviations in individual cases around the mean. Buying into the entire set of stocks has you achieve the central mean/average. Broadly, on average, sampling is considered as taking on additional uncompensated (on average) risk.