No. of Recommendations: 17
Cap weight was designed on purpose so that there would be minimal trading, not because it's the optimal way to invest for the investor with capital at risk. As prices go up, there isn't any trading to do, so it's cheap and easy to run a fund. That's the reasoning.
There is another important and rational reasoning besides the reduced trading costs: Market cap weighting distributes your capital exceeding more equally between each constituent income source within all firms across your investment, compared to equal weight indexing.
To explain, imagine two firms. Firm A owns 10 houses, and firm B owns 1000 houses. Let's assume all houses are largely indistinguishable and on average pay the same rent and are thus also equal in value. Ideally - thinking like a business owner - you want to be invested equally across all the houses. (For simplicity we consider assets but in the general principle applies equally by substituting houses with 'sales channel', 'dollar of earnings' or even 'customer' - as the business as a whole doubles in economic size we want to be investing double the amount.)
With market cap weighting you have 100 times more capital places in firm A but the same amount of capital across each house.
With equal weight investing, you are placing 100 times more capital in all the houses owned by firm A compared to firm B, despite the houses being indistinguishable. Furthermore you are almost not invested in firm B if your prism of view is on a per house basis.
Economically your investment is radically unequal with the equal weight index. Sorry to break the news given the attractiveness of the word 'equal'.
If you do want to invest (ie. expose) an equal amount of your own capital across each house, you would need to invest with a market cap weighting. That isn't perfect as some firms are overpriced and I will get to that, but it is relatively close to equal on a per house basis, rather than being out by two or more magnitudes which you get to when allocating equally merely on a stock symbol basis.
Despite this, equal weight indexing outperformED (uppercase added for emphasis) market cap weighting by about 1.5% the last 100 years. This must be reconciled however with the huge trading costs during most of the 20th C. Once the historical trading costs are added to the comparison, and tax is accounted for, some of the performance advantage with the equal weight index over most of the 20th C erodes.
Yet there is one good rationale for the outperformance and it is the one I referenced in the Manlobbi's Descent book - it continually forces capital away from firms temporarily overpriced into firms underpriced, especially useful when they are equally competitively.
(The other genuine equal-weight pro is moving capital away from market saturated sales lines into sales lines with a larger runway ahead. However that equal-weight pro is offset by the following equal-weight con: The smaller businesses are on the whole less competitive (try starting a small sugar drink firm) than large firms with their various network effects and the effect of lobbying/bribery (the same word, but difficult for small firms to do). So this it may be convenient for these pros and cons to be considered as exactly cancelling each other out such that we are left with the earlier mentioned equal-weight indexing pro - continually allocating capital (and sometimes as much as half the entire value of a firm such as when it doubles in price from popularity alone without any change to the business) from temporary overpriced situations into temporarily underpriced situations.)
There is a theoretically (but not practically) lower risk with equal weight investing - if - you frame the argument as a market speculator might think: You won't lose suddenly on a large position if just that position falls in price much more than the market as a whole. But - thinking like a business owner the converse risk is higher - the larger risk is with the equal weight indexing because of the exceedingly high concentration of you capital on a 'dollar per project' basis in the very smallest firms.
In truth the risk is not importantly different between equal weight and market cap weight index funds and what matters more to true investors is the relative long term value accumulation. The answer to that isn't trivially conclusive, once trading and taxation costs are added depending on where you live and how the fund is structured.
But owing to the 'overvalued capital moving to undervalued capital' argument alone, and less upon the long term backtesting showing a small advantage to equal weight (as that may not repeat with our present conditions of very low trading costs - akin to trading like Ben Graham no longer working) I do marginally favor the equal weight indexing over market cap weight.
- Manlobbi
