Bonds Down, Stocks Up!

So bonds are 'crashing' and stocks are going up. Some say that this can't continue; you can't have stocks and bonds go in opposite directions for long before something snaps. Well, this is true to some extent.

But it sort of depends on which way the 'snap' is happening.  What if bonds were way overvalued versus stocks? In that case, bonds can tank a lot before stocks have to correct. Bears have been saying forever that once the bond market turns and interest rates start to go up, the stock market will crash.

Maybe so.

Let's take a step back and look at what's going on. How can bonds go down and stocks go up?

Here is a look at the earnings yield versus 10-year bond yields since 1871.  The data is from Shiller's website:

S&P 500 Earnings Yield versus 10-year Bond Yield
(1871-November 2016)

Earnings yield and bond yields have sort of tracked each other closely since the 1960's or so. Recently, bond yields went down a lot while earnings yield refused to follow it down. This spread is sort of a cushion for the stock market. Since the spread is so wide, it's not unnatural for both of them to go in opposite directions.

Here is a close-up of this chart from 1950. You can see that that bond yield and earnings yield do sort of track each other, and there is a cushion between them so rising bond yields at the moment do not pose an imminent threat to stock valuations.

S&P 500 Earnings Yield versus 10-year Bond Yield

(1950-November 2016)

Back in 1987 and 2000, for example, bond yields were higher than earnings yield. Back in 1987 before the crash, this spread blew out from 1.3% in December 1986 to 4.4% in September 1987.  That was the rubber-band stretch that caused the market to 'snap' (Black Monday) back into normalcy.

In January 2000, the gap also blew out to 3.3% from less than 1% after the Asian Contagion of 1997.

Here is the spread between bond yields and earnings yields since 1950:

Bond Yield - Earnings Yield

(1950-November 2016)

...and here's a close-up since 1980:

Bond Yield - Earnings Yield

(1980-November 2016)

You can see that pre-Black-Monday-spike in the spread in 1987. Just because the spread is wide doesn't mean a correction is coming, of course. The spread widened many times since 1980.

Nor does it mean that the market can't correct or crash when this is negative. In fact, the spread was negative before the financial crisis in 2007.  That spike you see in the spread chart above actually happened in 2009 when earnings plummeted.

But it is sort of a big-picture-rubber-band indicator. A correction in bonds doesn't automatically mean we must correct in stocks.  One way to look at it may be that if the spread widens too much above zero, it's a warning sign.  We are far from that at this point. 

P/E's Too High

And since I keep hearing, and have kept hearing for years and years that the stock market P/E is too high so therefore must crash or correct, I created this chart to show how meaningless the comment about P/E ratios are when looked at alone.

From the above, we can sort of see that earnings yields and bond yields track each other.  I know many people argue that this Fed model is no good. OK, so maybe it's no good, whatever that means. But you can't really separate the risk free interest rate from asset valuations. Whether you want to use 2.3% to discount long term assets or not is another issue. But over time, of course the treasury rate is going to be a huge factor in determining asset prices.

So, instead of looking at yields, I just drew the same charts as above but using a P/E ratio.  Instead of a bond yield, I created the bond P/E ratio (inverse of yield).

Check it out:

S&P 500 P/E versus 10-year Bond P/E

(1871-November 2016)

...and here's a close up:

S&P 500 P/E versus 10-year Bond P/E

(1950-November 2016)

So you see how bonds are way more overvalued than stocks.  Well, actually, I really don't know if bonds are overvalued or not. In order to know that, I would have to know what future inflation and economic growth is going to be, and I don't know that.

I am always baffled at comments like, "The market has averaged a P/E ratio of 14x for the last 100 years so the stock market is 40% overvalued at 20x...".

How can you compare 14x P/E to the current level without discussing interest rates?  And if you think stocks should trade at 14x P/E today, then you should also think that interest rates should be much higher than they are now. For example, the 10-year bond rate averaged 4.6% since 1871 and 5.8% since 1950. But these periods include a time when interest rates were not set by the market.

OK. So far, we have determined that the bond market rout up to now is no cause for concern for the stock market. Yields of 2.3% is still way below earnings yield.

So What Should 10-year Treasuries Yield?
Obviously, this is the next question. I am no economist so I actually have no idea, but one idea I have always liked is that someone said that long term interest rates should be equal to real GDP growth rate plus the inflation rate, or more simply, nominal GDP growth. This sort of makes sense.

So below is a chart of nominal GDP growth versus the 10-year bond yield. The data is from the FRED (St. Louis Fed) website.

Nominal GDP Growth versus 10-year Treasury Rate

(1930-2015)

Not so bad tracking. Here's a close-up from 1950:

Nominal GDP Growth versus 10-year Treasury Rate

(1950-2015)

So not bad.

From this chart, you can see that the bond market is in fact overvalued, even with a yield of more than 2%. Either that, or the market is expecting nominal GDP growth of only 2.3% or so as of now. As usual, we don't know who is right.

I read a quote of Jeffrey Gundlach of Doubleline saying that he thinks long term rates can get up to 6% in a few years. I don't know if he was referring to the 10-year or 30-year. But who looks at the 30-year anyway, right?

But for yields to get to 6%, Gundlach must be expecting much higher inflation. Real GDP is probably not going to grow that much more than 2%.  Maybe 3%.  But if it does grow 3%, we only need 3% inflation to get to a 6% bond yield.  OK. Maybe that's possible.

So should we be worried about 6% long term rates?

Here is a snip from my Scary Chart post from this summer. 

1955-2014:

            Interest rate range           average P/E

                   4 - 6%                             23.3x
                   6 - 8%                             19.6x 

I looked at the data from 1955-2014 (adding one more year to update this isn't going to change much) to see what the average P/E ratios were when interest rates were in certain ranges.

From the above, we see that the market traded at an average P/E of 23.3x when interest rates were between 4% and 6%.  The 10-year now is at 2.3%.  So we have a long, long way to go for interest rates to threaten the stock market, at least in terms of the bond-yield/earnings-yield model.

Even if rates got up to the 6-8% range, the average P/E has been 20x P/E, or where the stock market is now on a current year basis.

So even if interest rates popped up to 6-8%, the stock market has no need for a valuation adjustment.

Of course, the market can still react negatively to big moves in the bond market, and of course, higher interest rates will impact earnings of companies with debt. This may be offset by a stronger economy if that's what leads to higher rates, not to mention higher earnings at banks and other financials that have been suffering under this low rate environment. It's always tough to model this stuff out.

As we have seen recently with the election, nobody really knows what's going to happen. And even when things are predicted correctly, the market reaction tends to surprise.

From the above blog post, here is a regression analysis of bond yields versus earnings yields. The 1980-2007 period seems overfit for sure; it's a period when stock yields and bond yields tracked each other very closely. I think I used the excuse that post-2007, we have been living in a post-crisis environment of unnaturally low interest rates.

So you can reject that regression as not valid.

Here is the original post when I looked at the relationship between bond yields and earnings yields:
scatterplot.

Again, keep in mind that long term rates after the post-election 'plunge' is 2.3%. The above regression shows that even at 4%  bond yields, almost double the current level, the market tends to trade at anywhere from 19x to 31x P/E.

Conclusion
So bond prices have tanked and the stock market doesn't care at this point. Bears say this can't go on and that overvalued stocks will follow bonds down soon.

The above shows that

• The recent plunge in bonds is a rubber band snapping back (bonds way overvalued versus stocks) and not a rubber band stretching that will eventually snap (like the bond crash before Black Monday).
• Bond yields are still below earnings yields so current bond market correction shouldn't be a threat to stock prices at current or much higher levels. 
• Bond yields will probably have to get much higher than earnings yield before it becomes a serious threat to stock prices. 
• Even with bond yields at 6-8%, stocks prices can be reasonably valued at current levels. 
• Saying the stock market is overvalued with respect to historical averages is nonsense without reference to historic relationship to bond yields.
• etc...

Having said all that, of course the market can still tank for any number of reasons; geopolitical issues, economy tanking, some sort of crisis etc. 

This only isolates and looks at two factors. And according to those two factors, there is no tension that needs to be resolved. 

Since I tend to post these and other charts on occasion, and it's a hassle for me to update every time, I am going to set up a companion website to this blog that will have this data and charts that I can update and reference so I don't need to update and cut/paste into a blog post. Also, anyone can go there and look at it at any time to see what these indicators are saying. 

There will be other information/data posted there too as sort of my public, personal investment notebook.  

I will provide a link to it when it's ready.