Last year I ran across my childhood stamp collection, including some mementos of the Weimar hyperinflation.
In a recent trip to Germany, I affixed the 10 Billion Mark stamp to my rental car and presented it at the local post office, asking them to kindly deliver the blue Opel wagon to the moon. They refused -- even though I offered to let them keep the change. An animated exchange of hand gestures suggested that even geo-synchronous orbit would be out of the question. Too bad the Deutsche Post doesn't honor their obligations. I had real plans for the other 2 Trillion worth of stamps in my collection. Like mailing Neuschwanstein to my sister in Portland.
Beyond Avis adding my unused LDW to their EBITDA, how much does all of this have to do with investing? Well quite a lot actually. Although I don't see hyperinflation in America's future, the Weimar experience tells us a lot about using monetary policy to tweak an economy. It has profound implications for future Fed policy and how investors should position themselves. So let's review what happened.
Before Ben, there was "Helicopter Havenstein"
In 1919, Reichsbank President Rudolf Havenstein started printing money to help the Weimar government cover its enormous deficits. Along the way, something interesting happened. The war-weakened economy got stronger and unemployment dropped. In fact, it dropped a lot, plunging to 2.5% after a year and a miniscule 1.5% after 3 years. Havenstein had unwittingly discovered the "Phillips Curve" -- to reduce unemployment, you just generate some inflation.
German politicians like Hugo Stinnes caught on to the idea. Stinnes claimed that inflation was "the means of guaranteeing full employment, not as something desirable but simply as the only course open to a benevolent government." So there was enormous political pressure to print money -- not that the Reichsbank really needed any extra motivation.
The hitch was that a spurt of inflation only reduced unemployment for a while. As the graph shows, when the rate of increase slowed in May 1920, unemployment immediately began to climb. So the Reichsbank increased inflation again. This worked until late 1921, when unemployment started to creep up again. It was no longer enough to increase inflation, the bank had to increase it at an increasing rate. Again, this only worked for a while, since people began to anticipate the increasing rate. So now the bank had to increase the rate of increase at an increasing rate (!!) to reduce employment. The rest, of course, is history.
Rational Expectations theory provides a nice explanation for this. Briefly put, inflation has no effect on the real economy if everyone expects it. When producers know that higher prices just reflect more money (rather than higher real demand), they just slap higher price tags on everything, rather than increasing output. When this happens across an entire economy, prices go up but real output and employment remain unchanged. For inflation to work, you have to trick people into thinking it's a demand increase rather than just inflated prices from too much currency. You have to stay ahead of inflationary expectations.
That's how the Weimar inflation worked so nicely at first. Everybody mistook all that new money for real demand. So they started producing more and hiring more workers. As it dawned on them that there was no real demand increase -- just more currency -- they started to raise prices and cut back production. It got progressively harder to trick them. With each go-around, the government needed to print even more money than was expected. But the expectations kept growing faster and faster.
The Reichsbank had many advantages that made it easier to trick people -- the press was tightly controlled and exchange rates were essentially a state secret. There was also heavy propaganda claiming that inflation was non-existent. Havenstein himself famously claimed that he was waiting for prices to go down to buy a new suit. Despite these efforts, people did eventually figure out that the Reichsbank was printing money at an accelerating rate and responded accordingly. Of course smart, rich people like Hugo Stinnes figured it out first and became fabulously wealthy as a result.
Even when average people anticipated inflation, they couldn't respond efficiently because they had to deal with actually paper money, rather than electronic balances that could easily be moved. On payday, workers would distribute bundles of money to relatives who would try to buy things as quickly as possible before the money became worthless.
Phillips "Curve" or Just an Orthogonal Line?
Despite the failure of the inflation-unemployment trade-off in Germany, in 1958, Kiwi economist William Phillips proposed the "Phillips Curve" based on a short period of English monetary history. Central banks, and particularly the U.S. Fed, quickly latched on to the concept. The idea that they could actually control employment must have been very enticing to them. The problem was, by 1970 the Phillips trade-off didn't work -- at all. This is particularly clear when you look at payroll data, rather than fuzzier unemployment survey, which omits discouraged workers.
(click to enlarge)
As the graph shows, each surge in inflation (the blue line, culminating in the green circles) is accompanied by a big drop in payroll growth (red line). If anything, this suggests the precise opposite of the Phillips curve: to kill job growth, you just generate some inflation. This is confirmed by the experience of countless other countries, from Brazil to Zimbabwe.
This evidence was lost on Congress, which, in 1977, gave the Fed its "dual mandate" to promote stable prices AND maximum employment -- as if the Fed actually has control over the latter. The Fed has nevertheless embraced the dual mandate, with numerous "Quantitative Easing" efforts aimed at increasing employment.
So how well has QE really worked? To find out, we need to look past unemployment survey data which are clouded by the record numbers of people leaving the workforce altogether. The percentage of people actually working (the employment ratio) gives a much clearer picture of what's going on. The graph below shows the employment ratio. Readers can decide for themselves whether the Fed's QE and "Operation Twist" (OT) efforts have had any positive effect on employment.
"But wait!" some savvy readers will exclaim, "The Fed hasn't even generated any inflation yet." Those readers are absolutely correct. They might not be happy about the reasons though.
Rational Expectations Evolve Again
Looking at the history of Weimar inflation, we can see how consumer expectations evolved to anticipate monetary expansions. The last 5 years of Fed policy (and particularly the new "transparency" inaugurated under Bernanke) have led to another evolution in expectations: banks are now anticipating and frontrunning the Fed. This has rendered the Fed even more impotent than before: the broader economy doesn't even get the chance to ignore the Fed's massive money printing, because the money isn't even getting that far.
Imagine you're a bank and you hear the Fed declare: "as long as the economy is slow and unemployment high, we're going to keep buying Treasuries and agency-backed mortgages month after month. When the economy strengthens, we'll taper and eventually reverse those purchases." What would you do? If you wanted to maximize your risk-adjusted return, you would borrow lots of Fed money at 0.25% and pile it into Treasuries and agency-backed mortgages (and maybe your proprietary trading desk) until it looked like the economy was about to improve. Then you would start dumping these highly liquid securities just before the Fed started to do the same.
What you wouldn't do is make a bunch of illiquid loans to businesses, consumers, and non-conforming mortgagors at low rates. You don't want to get stuck holding hard-to-unload 3.5% loans when rates go back up. That's why we've seen the percentage of agency-backed loans soar from 65% to 98% in the last six years. When it comes to mortgages, "non-conforming" is another way to say "non-existent." The percentage of all lending outside of government (or government-backed) debt has likewise shriveled. The end result is money velocity has declined to record lows and very little of the Fed's massive expansion has gotten to the public.
With velocity plummeting, the amount of money supply in the hands of the public (as well as inflation) has barely budged. So even if there were a Phillips curve, it has never even gotten the chance to prove itself.
The "Muppet" of Last Resort
The clear lesson from the Weimar Republic was that a central bank ultimately has little control over the real economy. The only way it can affect employment and output decisions is by fooling people about what it's actually doing. Back in 1920, with a controlled press and paper-based financial systems, you had a chance of fooling people -- at least for a while. Today, with instant 24/7 news, electronic finance, and the Fed telegraphing its every move, there is little chance of fooling anyone for very long.
On the contrary, with Bernanke's new transparency, everyone is positioning themselves ahead of the Fed. As QE efforts were announced, banks were stocking up on Treasuries, with their holdings reaching an all-time high earlier this year. As the "taper talk" began, they started selling, with the 10-year treasury yield rising by 80% in only 4 months, the largest percentage increase in history. Banks and other financial institutions got in ahead of the Fed and are determined to get out before the Fed does. And the Fed has made it all too easy for them.
What Should Investors Do?
Unless the goal was to pump up bank earnings or reduce labor force participation to the lowest rate in 35 years, the Fed's QE program has clearly failed. As poor employment reports continue to pour in, reflecting yet further drops in labor force participation, the question is will the Fed acknowledge the failure and abandon the policy?
My guess is that's what the current taper talk is really about: "let's point to the deceptively low headline unemployment numbers and use them as an excuse to declare victory and get out." It's a bit more comfortable than declaring we failed and getting out. If the current favorite, Larry Summers, wins the Fed chair nomination, it's a fair bet that he will further hasten the end of QE. By his record, he has been much less of a faith-based economist than the current Fed leadership.
Even barring that, at some point either the economy and employment will start seriously growing by themselves or the Fed will have to acknowledge the failure of QE. If I had to bet, I would say one or the other should happen within a year. At that point, we can expect interest rates to be substantially higher. We can be sure that banks will have dumped their massive Treasury and MBS holdings long before the Fed dumps its own.
For my part, I intend to continue front-running both the Fed and the banks by avoiding Treasuries and Agency bonds. I will also use any significant decline in rates to build a long-term short position. I would consider shorting the long-term Treasury ETF (TLT). For reasons, I've discussed in another article, I would avoid using the UltraShort Treasury ETF (TBT) to short Treasuries.
Additional disclosure: I am long Treasury Futures Puts.