Thursday, March 2, 2017

Buffett on Valuation

Buffett was on CNBC the other day and it was very interesting as usual. Well, most of what he says is not new.

Not in Bubble Territory
Anyway, he was asked about the stock market and since so many people keep saying that the stock market is overvalued or that it's in a bubble, I found it interesting that he says that, "we are not in a bubble territory or anything of the sort now."

He said that:
  • it's a "terrible mistake if you stay out of a game because you think you can pick a better time to enter it...", or something like that. He's been saying the same thing for years. 
  • if interest rates were 7 or 8%, prices will look "exceptionally high", but that measured against interest rates, stocks actually are on the cheap side.
  • if interest rates stayed at 2.30% over the next ten years, "you would regret not owning stocks".
  • we have to measure against interest rates.  Interest rates acts as gravity for valuations.
  • compared the long bond to an entity trading at 40x earnings with no growth and said stocks are attractive compared to that. 

This agrees with the charts/analysis I've been posting here relating P/E ratios with interest rates.

I know  there are quants who say this is wrong, that P/E ratios can't be compared to interest rates (we discussed this in the comments section of one of my posts about P/E's). I understand that argument, but history shows that the market does in fact use interest rates to value the stock market however theoretically wrong it may be, and the greatest investor of all time does so too.

What if?
OK, many commentators and pundits are baffled at this huge rally in the stock market thinking it's insane and taking the market to extreme valuations.  I have been posting here for a while that the valuations aren't all that extreme given the interest rate environment, even if you assume interest rates go up a lot from here.

Well, Buffett says if rates get up to 7-8%, then stocks are really expensive. But will rates get up that high? Even though I'm not an economist and have no idea what interest rates will do (actually there is no relationship between the two), I tend to believe that interests rates will get to 4-5% at most, on average.

So, let's play a simple what if game. This is not a prediction or anything, just a scenario that sounds plausible and is not at all a stretch.

What if GDP growth is stuck, more or less, at the 2% level?  Maybe Trump gets it up to 3%, but a lot of people don't think that is possible (except Jamie Dimon who thinks we can go much higher in terms of growth). And let's say that inflation does tend towards the 2% level.  That gets us to a nominal GDP growth of 4% or so over time.

Let's also assume that long term rates do revert to the level of nominal GDP growth.  Then long term rates would get to 4%. Of course, there will be overshoots in both inflation, GDP growth and interest rates. But over time, it's not hard to imagine interest rates averaging 4%. Total debt levels, demographics etc. make it hard to imagine higher nominal GDP growth.

So using the earnings yield-bond yield model, let's assume that the earnings yield tracks closely to the long term interest rate of 4%.  That means, over time, that the P/E ratio can average 25x in this environment.

Right now, 25x P/E ratio just seems super-expensive to many people because they look at the past 100 years and the stock market hasn't stayed at the 25x level for very long and has more often signaled a major market top than anything else.

But given the above scenario, it's not really inconceivable that the market P/E gets up to this level for an extended period of time. Some will argue that Japan has had lower interest rates and has been unable to sustain a high P/E ratio, but Japan has a lot of problems at the micro level too (companies not allowed to cut costs in their system of "corporate socialism" where large corporations are expected by the government to carry the burden of unproductive, unnecessary workers).


S&P 500 at 3250, DJIA at 29,000
The consensus EPS estimate for the S&P 500 index is $130. OK, I know that this will come down throughout the year, but that's what we have now so let's just use it. I'm not making a prediction or anything, just conducting a simple thought experiment.

Using the above, future average P/E of 25x, that would put the S&P 500 index at 3250!.  Using the same percentage increase, that would take the DJIA to 29,000!

Believe me, that sounds just as stupid to me as it does to you. I'm just making simple assumptions and plugging in numbers. My own personal experience (anchoring?), however, makes these figures hard to swallow.

But you see, it doesn't take much for the market to get up so high, and I am using a 4% interest rate, not 2.3%! So a large increase in interest rates is already built in.

Sure, inflation can get out of control and rates can go higher. I am just trying to figure out a long term, stable-state, through-the-cycle sort of scenario, and 4% rates and 25x P/E ratio just seems normal in that sense.  OK, so we can expand that to 5% rates and 20x PE; so let's just say rates can get to 4-5% and P/E ratios to 20-25x without it being stretched in any way.

Again, this sounds crazy and sort of feels like 'new era' thinking and Irving Fisher's "permanently high plateau". I know. Every time I make a post like this, I feel like I am putting in the top. But this doesn't feel like justifying anything new. In fact, I am insisting that things will go back to the way they always were; P/E ratios will be driven by interest rates, interest rates will be determined by nominal GDP growth rate etc.

I'm not making any outrageous assumptions like real GDP growing 4%/year or earnings growing 15%/year into perpetuity or anything like that.

And keep in mind that in this scenario, this is just the future 'average'. The markets can get much higher than 25x P/E in a manic phase and much lower in times of panic.

In fact, this has already been happening. The stock market has been overvalued in the eyes of many since the 1990's and hasn't reverted back to 'normal' levels in a long time. I think the error is that many look at raw P/E's and don't account for interest rates.

Not to be Bullish
And by the way, I have been saying this sort of thing over the past few years not because I am bullish; I am actually an agnostic (but bullish over the long term). I say this stuff to counter a lot of the "market is overvalued so it must go down!" argument and to caution people (and myself) to stay the course and act rationally.

Some funds claim to use a lot of sophisticated models and they write great reports, but at the end of the day, they are just net short the market (and have been for years!). That's just gambling; betting all their client's money on a single trade. Crazy.

Trust me, I get the same queasy feeling you do when I type 25x P/E. I honestly don't know what I would do with the S&P 500 index at 3000.  I know I would be very uncomfortable (if it happened within the next year or two).

So I'm not really being a Pollyanna.

When considering this stuff, it becomes less of a mystery why Buffett would spend $20 billion buying stocks since the election (or including some buys just before). And it becomes a big mystery why anyone would want to be net short this market (unless you are a short term trader who will be in and out as markets rally, like some hedge funds do etc...).

Sanity Check
And by the way, when all this talk of high P/E's make you nervous, just go check out my valuation sanity check page at the Brooklyn Investor website. This is updated after the close every day.

    Sanity Check

I often look here to make sure I am not seeing the trees looking like the Nifty Fifty.  When I see 20x or 30x P/E ratios on the S&P 500 index, I look at individual stocks to see if I see the same thing. If I do, maybe I worry. If I don't, I don't worry at all and assume the high market P/E is due to large cap, speculative names trading at high P/E's and/or hard-hit industries dragging down the 'E', or some of both.

Speaking of the Nifty Fifty, in the Bogle book I mentioned here the other day (Bogle book), he mentions a Jeremy Siegel study that showed that 50 nifty stocks bought at the start of 1971, near the peak, marginally outperformed the market over the subsequent 25 years.  Nifty Fifty returned 12.4%/year versus 11.7%/year for the stock market.

That's kind of crazy.  Even if you bought the Nifty Fifty at near the top, you would've beaten the market over the next 25 years, returning an above average 12.4%/year. By average, I mean the market returned 10%/year in the past 100 years or so.

Pzena Q4 Commentary
Pzena Investments posted their Q4 commentary and it follows up on their theme of the value cycle, and it is very interesting.

Check it out here.

Anyway, it shows that value has started to outperform again but that we are still early in the value cycle. Check out the tables and charts below.






I thought that was really interesting and I tend to agree with it. As much as I agree with Bogle and Buffett about indexing, there does seem to be a big, extended move in that direction which would have impact on valuations of individual securities, so it seems to make sense that maybe individual stock pickers can start beating indexes again (but don't bet on many being able to do so).

I still have a lot to say (or at least think about) in terms of fund managers, but that will have to be in a future post.