Or, "Why Inflationistas and Deflationistas Are Both Right"
Or, "Rethinking Gibson's Paradox (Part II)"
Or, "Should the Taylor Rule Go The Way of the Fed Model?"
I have more or less made my macro-calls for the rest of the year, so now seems like a good time to take a step back and look at the big picture.
Ever since the Federal Reserve and the federal government went into action to prop up the market, there has been a raging debate between inflationistas and deflationistas as to how the markets would respond. Inflationistas argued that negative interest rates and QE and bail-outs would result in inflation (or even hyperinflation) and high interest rates, while deflationistas have insisted that the debt-to-growth ratio is so high that the centralized actions to date are insufficient to prop up the markets. Since then, each side has had occasion to claim vindication. In early 2011, commodity prices were on a tear, with the inflationary bellwether of silver revisiting its old 1980 peak in a parabolic move upwards; at the same time, however, interest rates were skyrocketing in the European periphery. It soon became clear to nearly everyone (whether they wanted to know or not) that the eurozone was in deep trouble, and this resulted in a rush from commodities and stocks to safe havens such as treasuries and gold. So, now it is the deflationistas' turn to say "I told you so".
I'm going to try to explain why I think both sides are right but that it depends on how one defines "inflation". This is not merely a semantic debate, however. It requires us to expand our definition of "inflation" beyond price indices and year-on-year changes in them, but by doing so and differentiating between cpi and an understanding of inflation tied to the old form of Gibson's Paradox, this might direct us to a better understanding of markets and both how to make money in them and how to reform them.
I might be accused here of moving the goal posts to make the inflationistas happy, but I think they will have to accept a few things that some of them are reluctant to grant, as well. For one, cpi is not a fraud, and even if it is, it is not a consequential one. Second, there is no evidence of significant manipulation of precious metals markets.
Let's start with Gibson's Paradox and particularly the Barsky-Summers model of it. I have talked about this before, so I won't go into an extended discussion of it, but I would like to build on what I believe was accomplished before.
Under the Barsky-Summers model, metals prices were a function of low real interest rates so that low real interest rates resulted in high metals prices (particularly gold and copper). Interestingly, the real prices of those metals matched the level of inverted real rates. If real rates moved up and yet were still negative, metals prices tended to move down.
Unfortunately, time has been somewhat unkind to this model. There is an apparent relationship, but it is not especially clean.
If you look, it seems as if gold responds more positively to negative rates while copper responds more negatively to positive rate. And, that indeed seems to be the case.
What makes this even more interesting, however, is when we shift the gold/copper ratio forward sixteen months.
Under Barsky-Summers, high real interest rates should drive gold lower relative to prices (i.e., general prices should go up). Oddly, however, this does not seem to be the case with copper. By almost any measure, in fact, gold outperformed every other commodity and the cpi, as well, during the 1970s and early 1980s. The only exception might be silver.
This is as odd an outcome as the original Gibson's Paradox. But, whatever it might mean, it seems to give us a way to model the relationship between prices and real interest rates.
But what about the real prices of commodities? If real prices are not functions of real interest rates, what are they the function of?
It appears that something like the Barsky-Summers model occurs again, but in this case, it is year-on-year changes in cpi that matches up with real prices of commodities. In the chart below, I compare yoy cpi with a basket of real commodity prices (namely, gold, silver, copper, and oil).
Unfortunately, that was from 1970-2002. After that, things got really weird.
It took me a while, but it finally dawned on me that this looked a lot like the breakdown of the so-called Fed model which contrasts treasury rates with equity yields.
Equity yields are inverse functions of PE ratios, and PE ratios are effectively mirror images of the Dow/gold ratio. If that is the case, then real commodity prices (commodity/cpi) should be inversely correlated with the Dow/gold ratio.
This is kind of cheating, I suppose, because gold is located on each side of the equation, but it is suggestive nevertheless that real commodity prices are not a function of real interest rates, as in Barsky-Summers, but of equity yields.
One way of "testing" this is to compare the cpi/Dow ratio to commodities priced in terms of gold. Things get a little messy here, because I included gold in the original basket of commodities, so in effect, we are comparing real commodity prices (silver, copper, and oil) with the inverted real price of the Dow.
In a way, this is not surprising. High commodity prices are not good for the economy or for stocks, but on a short-term basis, that has not been the case over the last few years. Stock rallies have coincided with commodity rallies, but the stock rallies are increasingly feeble. This is not unlike the 1970s, but in our situation, this drag produced by commodity prices is not showing up in cpi.
In other words, whether or not we have textbook inflation, we are in the middle of stagflation: high commodity prices, high unemployment, weak stock markets, weak growth. And, yet cpi has been extremely low and falling for years.
Is it possible that the Fed should be targeting absolute price levels rather than rates of increase in those levels? Or, should it be targeting equity yields, which apparently is a rough equivalent of the same? I hardly have the capacity to hold an opinion on that question at this time, but if the Fed should be targeting absolute price levels, does that really give the Fed much to do, since under the gold standard, interest rates did that without compulsion?
In the meantime, it seems highly unlikely for the stock market to make much traction without a massive retreat in commodity prices, and yet commodity prices are now highly correlated with stocks, so the measures it would take to reduce commodity prices would, one imagines, be economically painful and politically unpalatable.
In this article, I have looked at commodity prices in bulk, but in my next article, I'd like to look again at movements within the commodity complex to gain more clues as to if and when this log jam will come unstuck.
Below, I put an estimate of the spread between ten-year yields and S&P yields next to the real price of copper.
It seems to suggest that we're in the right neighborhood when it comes to reconfiguring Gibson's Paradox, but not quite there yet. It also seems to suggest that if interest rates were raised above equity yields, that a reduction in commodity prices could be brought about.
Stuck between low yields and high commodity prices, there are few palatable investment options, so it will likely take relatively tactical maneuvering to gain returns until this market finds a way out. As the market creeps back towards its post-collapse highs, shorting equities (DIA) becomes an increasingly attractive move.
Additional disclosure: I am long September WTI and short September S&P 500 futures, as well as AUDCHF AND AUDJPY.