There seems to be increasing anxiety about which way interest rates are going to go and what this might mean for the stock market (SPY,DIA,QQQ). In this article, I will argue that a) long-term Treasury yields will probably fall over the next few years, perhaps quite precipitously, b) this implies an increased chance of a temporary stock market crisis in two or three years, but c) stock and bond market (IEF,TENZ,TLT,GOVT,UST) performance will probably accelerate until inflation picks up and short-term rates rise.
Historical Relationship Between Stocks And Interest Rates
So, what is the relationship between interest rates and stock prices? If I were forced to provide a single unequivocal answer to that question, I would have to say that they are negatively correlated. Since 1970, the correlation between the ten-year yield and real stock prices is -0.74. Falling rates are, in that sense, "good" for stocks.
(Sources: Unless otherwise stated, all data comes from Robert Shiller)
Obviously, though, you can see there is more going on than what is suggested by that forty-year correlation. For example, from the 1940s until the 1960s, interest rates and stocks rose together. Or, looking at nominal stock prices over the last forty years, stocks have only been correlated with interest rates for any considerable time during the bear markets of the previous decade and of the late 1970s. In bull markets, they are more likely to be negatively correlated.
Therefore, even if we guess correctly which way Treasury yields will move, this by itself does not tell us much about how the stock market will react. Without knowing anything else about the market environment, falling yields are somewhat more likely to be bullish than bearish, but not much more could be said beyond that.
The Yield Curve And Crashes
There is one reason falling yields might be cause for concern, though. If the yield curve were to flatten considerably, so that the spread between ten-year and three-month Treasury yields were to approach 1.0, a stock market crash and/or recession within the following five or six quarters would not be altogether unlikely, especially if an oil shock occurred somewhere along the way. (Nearly every oil shock, measured per Stephen Leeb as an 80% year-on-year jump in crude prices, has been preceded by a significant flattening of the yield curve).
(Note: In this chart, the ten-year yield is subtracted from the three-month yield to illustrate the way in which a low yield curve spread leads oil shocks and recessions).
Considering the slack in the economy, a fall in the yield curve spread is unlikely to be generated from a rise in short-term yields, but could it be generated by a 150-basis point fall on the long end of the yield curve?
Wagging The Dog: When Long-Term Rates Control The Yield Curve
This has happened before. Twice, actually, and it almost happened again in 2012.
Click to enlarge (Source: Shiller and St. Louis Federal Reserve)
In 1986, the yield curve spread briefly touched 1.0 because of a fall in the long-term yield. That was also a period of rather severe "lowflation." Oil prices collapsed that year, and by December, inflation was at 1.1%.
A year later, however, oil prices jumped 80%, interest rates suddenly rose, inflation was picking up, and stocks tanked.
Another instance of the yield curve spread being driven down by the long end of the curve occurred about ten years later. From February 1995 until September 2001, the yield curve spread remained below 2.0, and in 2000, the curve was inverted because of rising short-term rates in a bid to check inflation and an oil price spike. But, that initial suppression of the spread (from 2.07 in January 1995 to 0.7 in June of that year) was almost entirely a function of the fall in the ten-year yield from 7.78 to 6.17.
From late 1997 to late 1999, the spread would fall below 1.0 again even as short-term rates bottomed. Commodity prices collapsed, and inflation was back below 2.0% yet again. A boom in stocks arguably began in December 1994 and did not run out of steam in any meaningful sense until the spike in oil in 1999, even though this was a period of crisis in Asia and emerging markets elsewhere around the globe.
So, we can elaborate on our basic statement that falling yields are "good" for stocks. Falling yields are good especially when we are in a bull market, but they can be dangerous if they should drive down the yield curve spread to an unusually low level, even if that level does not have a negative sign. On the bright side, it usually takes a year or more for the negative "consequences" to manifest themselves and there are often other warning signs along the way.
Unfortunately, that is still too simple a description of the relationship between falling interest rates and the stock market.
The Yield Curve And The Parabolic Risk
The problem is that not only do stocks and bonds tend to correlate with one another during bull markets, but the suppression of the yield curve caused by a fall in yields is generally marked by an acceleration in stock performance. In other words, when the fluctuations in the yield curve are determined by the long end, the stock market tends to outperform. That may turn out to be bad news after a year or two or three, but in the intervening time, parabolic performance in the stock market appears to be typical.
I tried to draw this relationship out in the following chart. I took the volatility of the ten-year yield relative to the volatility of the yield curve spread (i.e., the rolling, five-year beta of the ten-year yield) and compared it to the performance of the S&P 500. Whether the relationship is causal or even "real" is beyond the scope of this article. What is significant, rather, is what this means about what a bull market is supposed to look like. The notion that falling yields are bad for stocks is quite mistaken, and most especially in a low-yield environment such as the present one.
I would argue, therefore, that if we should see a significant decline in the ten-year yield from the recent tops, we should not only expect stocks to benefit, but to go parabolic. It should be pointed out, however, that this particular relationship suggests that the direction of the movements in the ten-year yield would not matter in a zero-interest rate policy environment, because the correlation with the Treasury spread will always approach 1.
Nevertheless, how likely is it that interest rates will decline significantly anytime soon? More likely than not.
Interest Rate Cycles Since 1970
There are two reasons why.
First, Treasury yields tend to be highly susceptible to cyclical behavior. Not only do they tend to peak every five years and trough at similar intervals, but they seem to experience 69-week cycles (which is itself broken down into three 23-week segments), as well.
Second, although Treasury yields' most important and complex connections appear to be to the earnings and dividend yields alongside inflation, the most important predictive relation is with the gold/oil ratio, which tends to lead interest rate movements by about 15-16 months (again, roughly 69 weeks).
All of these indicators appear to be pointing to a decline in yields.
If, as seems likely, long-term bonds bottomed earlier this year, we can expect a decline in yields by something like 200 basis points over the next three years. Five-year lows tend to be at least 200 basis points below their previous peaks and, since the 1980s, 100 basis points or so below their previous troughs. The previous trough, two years ago, was 1.53 (using monthly data). Even a half-hearted adherence to tradition would see long term yields at 1.0, leaving the yield curve spread very narrow.
If we look at the shorter-term interest rate cycles, such as the 69-week cycle, and if we mark the previous bottom in July 2012, then November 2013 would have been around the time of the subsequent top. One could expect the following five or six quarters to see a considerable decline in yields.
(Source: Shiller and World Bank)
The gold/oil ratio also peaked in the fall of 2012, suggesting a peak in yields no later than March 2014. The gold/oil ratio then declined about 45% into July 2013 and has remained quite low ever since. Although the magnitude of the moves in the ratio are not necessarily reliable indicators of interest rate movements and no such indicator is foolproof, the signal is clear. Counting off sixteen months, it would appear that yields will fall for the remainder of the year and then remain rather low going into 2015.
I tend to think that the risk over the course of the next couple years is to the upside: a powerful performance in the stock market, coinciding with weak inflation, falling interest rates, foreign troubles, and falling commodity prices (especially precious metals, although oil may get hit first). By the time the real economy starts to improve, foreign crises begin to abate, unemployment normalizes, capacity utilization rises, and inflation begins to perk up, the Fed will try to cool things down, and the parabola (call it a "bubble" if you prefer) will pop.
So, what could go wrong with this outlook? In broad terms, the greatest threat to my argument is a breakdown in the relationship between yield curve behavior and stock market performance. The relationship is relatively "new," insofar as it really only seems to have held since the 1960s, roughly the time when the Dollar Equilibrium was established. That means we can say with absolute certainty that this relationship is at worst outright false or at best only capturing one dimension of a more complex relationship.
What if interest rates rise? Obviously, that would be none too good for my argument that bonds are set to rise, and I would also then be slightly less optimistic about the likelihood of stocks going parabolic in the near term. But, because in bull markets, interest rates and stocks tend to move inversely, I would regard the likelihood of a near-term correction as somewhat increased while remaining quite bullish, and in fact, it would marginally raise my expectations as to the duration of the current bull market.
What would it take for both interest rates and stocks to collapse? This would be the Black Swan. Of course, if this were to happen, the bullish bond prediction would look much better than the bullishness on equities. This would, however, be worse than if both predictions went wrong. For this sort of thing to happen, a political or economic crisis of some magnitude would have to hit more or less out of the blue, a nuclear strike by a rogue state or something along those lines. Really, though, this is an argument more about the necessity of maintaining a suitable degree of diversification in a broad range of assets rather than a reason to avoid market predictions.
What if interest rates go sideways for the next three, six, twelve months? I would feel aggravated but not aggrieved. I would regard this is a delay rather than a reversal. Of course, we would also have that much more time to read the tea leaves and modify our positions.
Finally, how long will we have to wait before we know whether or not this thesis is right or wrong? This article is looking forward about two to three years, but I expect that we will see some degree of confirmation over the next twelve months.
For me, this is an examination of only a couple pieces of the market puzzle. More generally, my view of the stock market is that all signs are bullish, and until something happens that convinces me otherwise, I remain optimistic about the short- to medium-term (up to the next five years) regarding stocks if pessimistic about what that means for the long-term (the 2020s).
In sum, I think we are on the brink of a crisis, although one that is in many respects already here, but the crisis will be a melting-up in American stocks and bonds while commodities and inflation get crushed and the real economy still looks pallid, if not as violently ill as emerging markets. The risk, in other words, remains to the upside.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.