Also, you say: “What’s fascinating to learn is that if the asset has an expected return equal to the rest of the portfolio, including it in the portfolio would not only decrease volatility but probably increase return as well.” This is surely a good argument for having a diverse holding of index funds for different stock markets, rather than just one index fund for all of your equities. The only problem is that stocks are globally quite well correlated, I suppose.

Bernstein’s point–which I tried to simplify and surely distorted somewhat in the process–is that the return earned when rebalancing is still (more often than not) going to be greater than that earned by “letting it ride” (particularly when the asset classes involved are volatile, and/or have low correlation).

Assuming that the 8% and 4% returns over the next ten years are geometric (i.e. compounded) averages, would not the expected 7% return for the 75/25 portfolio be attained only assuming constant rebalancing? Surely you will agree that if you did not rebalance at all, but allowed the stocks (8%) to run you would do better than 7%? (I get 7.12%.)

I don’t think the other commodities would be a very good proxy for valuating gold pre 1970- they would correlate but not perfectly. Bonds have a long track record so predictions of long term rates of return should be fairly accurate. There should be commodities with long records so you could calculate dependable statistics for them. I don’t have those statistics, but my feeling is that investing in any single commodity has uncompensated risk that could be diversified away through investing in many different commodities.

I’m not sure if it would be practical for a small investor. There are some commodity ETFs, but what I would really want is something similar to a commodity index fund.

-Rick Francis

As to the question of whether gold is currently inflated or not, I guess my reply is “compared to what?”

With stocks (and bonds), there’s an inherent value that’s a function of the fact that the security should actually be earning a return (dividends or interest) for its holder. We can therefore compute whether we’re paying a lot or a little for each dollar of earnings at any point in time. That is, stocks have a value other than what you can get for them on the open market.

With gold, its only value is in what somebody is willing to pay for it.

It seems then, that asking if gold is currently inflated is simply the same as asking “is it priced higher now than it will be in the near future?” And to that, my answer is simply “I haven’t a clue.”

I like the idea of considering other commodities. Are you suggesting considering them for their own sake in a portfolio? Or simply to use historical data from other commodities as a proxy for historical data for gold?

If option #2, then: Do other commodity prices behave the same way that gold does (or, more to the point, “will”)? It seems at first glance to me that the cross-comparison between commodities is rather more of a stretch than a cross-comparison between stocks.

Sorry if I’m rambling…As I mentioned, I’m entirely inexperienced regarding commodities, so I’m simply typing as I think this through…

I looked at a graph of the yearly gold price data- and I have to ask is the current price of gold is inflated? The price has nearly doubled in the last four years! I suspect because people are fearful of stocks and have fled to gold- a trend that could reverse very quickly.

If the current market is a bubble then it would certainly skew the average return figures for gold, especially since we don’t have that many years of data.

>I just think that coming to the opposite conclusion–”bonds >outperform gold over long periods”–is similarly questionable.

Since the statistics for gold are complicated by the gold standard why not choose some other commodity? I’m sure there must be historical data for other commodities. Oil has been pretty volatile recently and I don’t know of any currency pegged to the price of oil. I suspect there are hundreds of years of data on corn prices- and it seems that it could be more volatile as the yearly production would vary with weather conditions.

-Rick Francis

Basically I just plugged year-by-year prices into excel, then calculated the return.

Also did not know that about the gold… where do you find that data? j/w..

You’re absolutely right: There must be a point at which an asset’s stand-alone return is low enough that the rebalancing bonus would not make up for it.

As to gold vs bonds though, I don’t think it’s a clear-cut answer.

When most people quote super-low long-term gold returns, they’re usually looking at a period dating back to 1900 or so. The problem there is that we were on the gold standard for the bulk of that time period (up until 1970 or so, I believe).

As a result, I’d argue that the period from 1900 all the way up to the point at which we dropped the gold standard offers roughly zero predictive value in terms of what we can expect in the future.

Over the period 1970-2008, gold prices increased at a rate of 8.75% per year. 10-year T-Bonds earned 8.04% over that period according to my data.

Of course, I’m not saying that we should necessarily expect gold to outperform bonds. (Its winning margin is pretty slim, and 39 years isn’t a terribly long period from which to draw conclusions.)

I just think that coming to the opposite conclusion–”bonds outperform gold over long periods”–is similarly questionable.