Gold's Strange Behavior Shows It's No Haven

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Scott S
Posts: 1926
Joined: Mon Nov 24, 2008 3:28 am
Location: building my position

Re: Gold's Strange Behavior Shows It's No Haven

Post by Scott S »

NiceUnparticularMan wrote: Thu May 12, 2022 10:34 pm
000 wrote: Thu May 12, 2022 9:17 pmThe banking system makes USD, not the Treasury
Well, it is true that it is the Federal Reserve which determines how many USD are created.
Is that even true when the reserve ratio is 0%?
"Old value investors never die, they just get their fix from rebalancing." -- vineviz
Posts: 987
Joined: Wed Apr 24, 2019 2:19 am

Re: Gold's Strange Behavior Shows It's No Haven

Post by invest4 »

jason2459 wrote: Tue May 10, 2022 10:40 am I'm not a gold bug by any stretch of the imagination but using just a couple weeks of data to come to any conclusion for anything is completely ridiculous and foolish. :oops:
+1 I don't own any gold and have no plans to do so, but fully agree.
Posts: 3822
Joined: Sat Mar 11, 2017 7:51 am

Re: Gold's Strange Behavior Shows It's No Haven

Post by NiceUnparticularMan »

ivgrivchuck wrote: Fri May 13, 2022 12:58 am Perhaps it's as you say and the shift in inflationary environment will change the correlation in more permanent way, we'll see... if that happens, I'll need to change my perception about commodities.
So to be clear, I am making no predictions about "permanent" changes. I am pointing out the correlation between stocks and CCFs depends on what is happening with unexpected inflation and stocks. I can't predict unexpected inflation and stocks from here, so I can't predict the correlation from here either.

But what we already know is CCFs once again behaved as expected when certain circumstances occurred. You can therefore use that in your planning for the future, or not as you see fit. But always, it should be planning based on forward-looking estimates of the nature, "If X happens, then CCFs will probably do Y." It should not be just "CCFs will probably do Y", when whether or not X happens is itself not predictable.

I note again the Vanguard white paper on unexpected inflation was very clear on this. You can't justify an investment in CCFs based on something like a risk model where short-term volatility is your only measure of risk. You can only justify it, maybe, if you are particularly concerned about tail scenarios involving unexpectedly high inflation.
If we look at gold and stocks correlation that you posted: ... &months=36

It has gone up and down like crazy, but those two assets haven't ever stayed strongly correlated for an extended period of time.
So I agree that because free-floating gold in particular has such a large speculative component, it becomes far less predictable even contingently. In every way, including as to correlation.

To continue the above, you also can't say, "Free-floating gold will probably do Y." But you further can't say, "If X happens, free-floating gold will probably do Y." You can't really saying anything about what free-floating gold will probably do, because it is subject to swings in large swings in real value that are not connected to anything else.

So then what use is free-floating gold for any sort of planning purposes?
In principle, I agree. However we need to make a clear distinction between 0% nominal return and 0% real return assets.

In the era of very low (even negative) real interest interest rates, the picture is less clear.
I agree there is a lack of transparency about what to expect in terms of real returns on free-floating gold given today's prices. But that is not really a point in its favor.

Generally, I am not aware of a good reason to believe free-floating gold would stay at 0% real if, say, short-term bond rates were well below 0% real.

Again, free-floating gold in inherently unpredictable due to its market price being based on greater-fool reasoning. But, generally speaking, if savings versus asset supply dynamics have systematically run up prices, and down real returns, on stocks, bonds, and so on, then it is a better bet the same thing has happened to free-floating gold than not.

This is a critical point about transparency that I think too often gets lost here. TIPS are very transparent about their expected real returns. Nominal bonds are only transparent about their expected nominal returns. Stocks are even less transparent--about all you have is various common valuation measures. And then something like free-floating gold doesn't even have that, because as a greater-fool asset its price is disconnected from normal value measures.

But that doesn't mean nominal bonds, stocks, free-floating gold, and so on are not subject to the same capital market dynamics as TIPS.

Unfortunately, though, some people have a tendency to fill in things like recent historic averages for the expected real returns of less transparent assets. And in times of historically high prices, that means systematically favoring less transparent assets just because they are less transparent. Which is not a good assumption to make.
For example, let's run an example assuming that inflation is 3% and real interest rates are zero:
Stocks: 7% nominal expected return (vol: 18%)
Bonds: 3% nominal expected return (vol: 6%)
Gold: 3% nominal expected return (vol: 24%)
So in this case, you have assumed free-floating gold has the same expected real return as intermediate-term Treasuries. Generally speaking, that is not really a supportable assumption given the fundamental economics involved, as there is no reason to expect a term premium for free-floating gold.

So if you were going to use anything from bond data for free-floating gold, it should probably be something like the 1 month Treasury rate, currently at around 0.6% nominal.
and assume zero correlation between all asset classes.
Which you shouldn't do either. Again, it is a variable, but the long-term average correlation between free-floating gold and stocks is positive. It has indeed spent more time in positive territory than at zero or negative. So you should at least fill in something mildly positive.

But anyway, if you just do the bit about returns, free-floating gold drops out of the efficient frontier by your way of estimating it: ... eight1=100

In other words, your version of this model depended on implicitly assuming a term premium for free-floating gold, when there is no good reason I am aware of to make such an assumption.
Now, I readily admit that the chosen assumptions were somewhat favorable to gold (e.g. correlation is not exactly zero), but the takeaway is that adding gold in the era of low real interest rates is unlikely to take you very far from the frontier (as long as you stay within 5%-10% of the portfolio), and there is no long term change in correlations... Is it worth it? That is hard to say.
OK, so assuming the high asset prices that are showing up transparently in TIPs, less transparently in nominal bonds, and less transparently still in stocks, also apply to free-floating gold, then the answer using ONLY this sort of analysis is no, it isn't worth it. But it is true given ONLY this sort of analysis, the cost of doing so appears low. So far.

But there is a different mode of analysis which also applies to free-floating gold, namely that as a greater-fool asset it is subject to speculative booms and busts, and these busts can take the form of real returns being far lower than expected over long periods. Or far higher. That is the nature of these greater-fool assets.

This sort of risk is not captured by "volatility" risk. This sort of long-term real returns risk is an entirely different sort of risk than volatility, but no less real. And yet there was nothing in your risk model that captured this sort of risk.

OK, so by adding free-floating gold to a stock/bond portfolio, given realistic/consistent pricing assumptions, there is no good reason to believe you are getting a benefit in the form of greater efficiency as to volatility risk. But there are very good reasons to believe you are taking on additional long-term real returns risk.

So why would you do that? Why expose yourself to the risk free-floating gold is currently in a speculative bubble that will burst and cause you large real losses, if you are not getting a significant offsetting benefit?

You'd really need a compelling reason to take on such long-term real returns risk. And when you properly model free-floating gold, including realistic/consistent assumptions about pricing, there is no such compelling reason that I know of.
Posts: 459
Joined: Wed Aug 21, 2013 7:12 pm

Re: Gold's Strange Behavior Shows It's No Haven

Post by Gaston »

halfnine wrote: Thu May 12, 2022 1:30 am
Gaston wrote: Wed May 11, 2022 7:24 pm
Fremdon Ferndock wrote: Tue May 10, 2022 10:10 am At the risk of stating the obvious, inflation rates are very high and the performance of gold has been terrible.
As Larry Swedroe and others have noted, gold does tend to keep pace with inflation over long periods of time. And when they say “long periods”, they are quick to point out that this means “over centuries”, not within a human life span.
I can't find anything factually accurate in your statement.
I’m not going to debate the merits of owning gold. If you wish to own gold, that is your decision.

I do, however, stand by my interpretation of Mr. Swedroe’s viewpoint, which he has expressed on multiple occasions. For example, to quote him directly rather than paraphrasing his comments, he has stated that “research by Claude Erb and Campbell Harvey showed that, after conversion into gold, the military rate of pay between U.S. Army captains of today and Roman centurions under Emperor Augustus is comparable. If your horizon is that long, it seems like gold would be a good inflation hedge, though it provided no real return over that 2,000 years.” (From ... 4_2014.pdf)

If one chooses to interpret the above as “a few decades”, then again, you are free to reach that conclusion.

Regardless, whatever you decide to invest in, I do wish you every success.
Post Reply