Sometimes the stock does better than the investor

[martingale]

http://www.nytimes.com/2007/05/03/busin ... nd&emc=rss

Over the last 80 years the US stock market returned an average of 10% but the imputed average US stock investor earned only 8.6%

[Norbert Schlenker]

To be fair, this looks like blind application of money in and money out. While some of that is market timing, a lot of it is "We didn't have the money way back in 1926 but dribbled it in - as we saved it - over the years at higher average prices." So it's not all buy high sell low, it's just buy when you have the money.

[martingale]

Sure, but for whatever reason the money went in, and for whatever reason the money went out, the implication is an 8.6% return for the average investor over the last 80 years rather than the 10% ordinarily quoted. They may well have been transiting in and out of the market to fund real world expenses, rather than market timing, but nevertheless, transited in and out and earned 8.6%

That has rather simple implications for people who plan portfolios based on expected returns. Unless you _know_ you can buy and hold for long periods, the 10% return is not a reasonable estimate for your US stock return.

(Yeah, it's not for other reasons as well, but whatever you think the long-run return of the STOCK is going to be, subtract a bit to arrive at what the average INVESTOR will get.)

[bubbalouie]

For once I agree with you MG; I think you're saying that the masses are stupid.

[patriot1]

I think there is more to it than that. One that comes to mind is that the time when investors are most likely to need to sell is often the worst time to sell. I.E. job loss in recession.

Also true for other assets like housing.

[sevimo]

I think that an example in the article is misleading without defining who is investor they are looking at? If it's anyone who holds shares at the moment, then average stock performance must match investors performance. In the example they have, long 100x$10.0, long 10x$20.0, short 20x$10 (=-1000$), there must be someone who did exactly the opposite to close these deals, i.e. short 100x$10.0, short 10x$20.0, long 20x$10 (=+1000$). If we count the counterparty as another investor, then the average gain/loss is -1000$ + +1000$ = 0$, exactly the same as for buy and hold investor. The question is just how counterparties split the stock market gains.

Now, both these examples ignore frictional costs (this will make market timing from zero-sum game in this example into negative sum - pure waste), and we can artificially ignore institutional investors/share buybacks/etc and concentrate on individual investors (aka suckers , who consistently buy high and sell low). But I think the real reason for lower returns is transaction costs. Basically, naive market timing probably increases volatility and hurts your expected returns due to frictional costs, all at the same time.

[IdOp]

Another confusion: the NYT article talks about

a value-weighted portfolio of stocks ... earned an average annual return of about 10 percent.

(my emphasis) while the paper abstract only mentions broad market returns (I haven't seen the whole paper). Obviously these two are not the same thing.

ADDED: Maybe he means market value weighted, rather than value investing, in which case my confusion is solved.