Tuesday, March 10, 2009

Gold... Has It Really Stored Value?

In response to yesterday's inflation adjusted S&P 500, we were asked:

Could you present the same plot but with inflation-adjusted gold overlayed on the data? It would be interesting to see if stocks really do outperform over the long-term...
Over the past 34 years, the price of gold is only 25% higher in real terms.

Compare this to yesterday's chart to equities... (note this is back one more year than the chart for gold... I can't dig up that last year of gold data), which shows equities returned 400% in real terms over the past 35 years.

And the following chart combining the two 10-year rolling periods...


You shouldn't be... asset inflation (whether via gold, housing, or commodities) serves no long-term productive purposes... only increased productivity creates actual economic value. Thus, the movement of capital towards productive resources (i.e. what equities, fixed income, and bank loans are intended to do) is the only type of long-term investment that really creates broad economic value.

Using an extreme example: Imagine if all the wealth in the world simply went to buy assets (to be more extreme, lets pretend it all went to gold and was in turn stored in a vault). There literally would be no global economy.


  1. That's good. I think the next logical step would be how far investment returns can diverge from real underlying economic growth (ie Real GDP growth).

    My inkling would be that when it diverges too far (credit bubble) it crashes and returns to the underlying growth. I don't know that you could do that analysis though, because it's a varying subset of participants that moves capital into the investment classes (ie you could see a bigger divergence if only 5% of the population was investing than if the whole population is investing).

  2. I read somewhere, and can't find again, that industrial/consumer demand for gold is down considerably and only investor demand is higher. This doesn't sound like a "store of value" to me. It's just another "thing" that you can't eat, burn or wear.

  3. Does the published S&P500 number represent a constant metric over time? To get a more accurate picture, you may need to compensate for the positive bias in the S&P500data that is caused by the fact that poor-performing companies are dropped from the index (I'm not sure if the SP500 already adjusts for that but I doubt it). And then there are the fees that retail investors pay: 1-3% annual fees deducted at the beginning of the year. How much are we ahead of gold after that?

    Is the market really returning profits to shareholders or just to the bankers, fund managers, and bondholders?

  4. it is actually pretty inexpensive to invest in the S&P passively. the same argument (or worse) could be made with gold. investing in gold is often higher fees or you take a brutal bid/ask on the physical stuff...

  5. Great post. One remark/hint: When using the 10yr rolling return it would be better to show an annualized figure (i.e. 10yr CAGR). Mika Kevo

  6. Is this the total return on SPX, or just the price return. That definitely matters over time. Last time I checked, gold didn't pay a dividend. However, it a block of it will hold a door open.

  7. w/ dividends is what total return is genius.

  8. I recently posted a long term inflation adjusted chart of the S&P 500 going back to the early 1900's but it didn't include dividends. Interesting to compare the two.

  9. Classical Economics will tell you EXACTLY what the "real" return of holding gold will be: ZERO. Because price indices revert to the gold price over long periods (see Roy Jastram's "The Golden Constant"), gold, in fact, IS the "reality" one should be using to disaggregate the "real" from the nominal, in real time.

    While holding gold will provide zero real return over long periods, its still a better bet in a period of aggressive inflation than assets denominated in the paper currency which is being debased.