A.I.M. Users Bulletin Board

[ls7550]

Re : Equal weighted

Couple of historical equal weighted index data

http://www.bogleheads.org/wiki/Equal_weighted_indices#Appendix

http://tinyurl.com/k8zbya6

With equal weighting, say 1% in each of 100 stocks, there's a finite downside per stock (100% loss of the 1% allocation), but a (theoretical) unlimited upside. 1% allocation to a 10 bagger (10 fold gain) adds 10% to the total portfolio value. 1% allocation that loses all hits the entire portfolio for a 1% decline.

A bit like buying lottery tickets. With equal weighting you select different numbers for each ticket, with cap weighted that perhaps has one stock 10% weighted its like buying 10 tickets out of the 100 and putting the same numbers on each of those ten tickets. Large stocks also have less tendency to gain 100's of percents, whilst smaller companies can/do. IMO that explains some of the 'small' premia. Value generally represents something that has already endured a sizeable decline and where that loss was carried by someone else, in effect reducing the downside some (broadly) for the buyer of value. Hence a small-value premia (tendency to relatively outperform). That's not a constant thing however as its more a lottery type thing (sporadic/intermittent). Generally if you look at the histories of such indexes for much of time they might broadly compare with the average, but its the less frequent big up outliers that generate the overall outperformance (occasional winning lottery ticket).