The Alpha and Omega of my Thoughts on Backwards Induction from the Long Run Omega Point
This extraordinarily long and confused post is my response to Brad DeLong’s
New Economic Thinking, Hicks-Hansen-Wicksell Macro, and Blocking the Back Propagation Induction-Unraveling from the Long Run Omega Point: The Honest Broker for the Week of May 31, 2015
Before the jump, I make only three brief points.
1) The phrase “the Long Run Omega Point” asserts that there is only one possible long run outcome. Backward induction from the long run requires that there is no hysteresis (that is path dependence — the word was pulled back from Larry Summers distant past by one Brad DeLong).
2) Someone smart once said something smart-assed about backward induction.
3) Just so it’s clear — I have never met anyone who I am convinced is definitely smarter than Brad DeLong
more thoughts (much much more) after the jump.
Brad’s argument (somewhat truncated)
Because the long run will come, increases now in the monetary base of sufficient magnitude that are believed* to be permanent will–maybe not now, but soon, and for the rest of our lives, in this long run–produce equal proportional increases in the price level, and thus substantial jumps in the inflation rate as the price level transits from its current to its long-run level.
Moreover, there is more to the argument: The long run is not here. The long run may not be coming soon. But the long run will come. And so there will will a time when the long run is near. At that time, those who are short long-term bonds will be about to make fortunes as interest rates normalize and long bond prices revert to normal valuation ratios. At that time, those who are leveraged and short nominal debt will be about to make fortunes as the real value of their debt is heavily eroded by the forthcoming jump in the price level .
And there is still another step in the argument: When the long run is near but not yet here–call it the late medium run–investors and speculators will smell the coffee.
etc to the medium run proper but not to the short run.
*In passing, when discussing what might happen given the current state (here the balance sheet of the Fed) and arguing that it depends on what people might, upon reflection, believe, it isn’t really kosher to add an assumption about their beliefs. But, for the sake of argument, I will assume that the increase is permanent and believed to be permanent.
The argument depends on the logic of Nash equilibrium. Agents must assume not only that they understand the story, but that many others do (not everyone but enough and as rich, so those who understand will determine asset prices). Similar arguments have been used to prove that there can’t be asset price bubbles. It is odd to find the argument in a discussion of the policy response to the collapse of a huge housing price bubble.
It only works if the average bond trader doesn’t think that he is smarter than the average bond trader (if so why the hell is he actively trading bonds ?). If the average trader thinks he can sell long term nominal bonds to a greater fool, when the time comes, then the backward induction fails.
Also, the logic is (mostly) deterministic. it doesn’t work if the end of the liquidity trap is triggered by a sunspot variable so that month after month there is a 1% chance that it ends this month.
The argument proves too much — there is no reason that the induction shouldn’t go back to the short run — say January 2009. If one can’t explain why interest rates and inflation breakevens didn’t spike immediately, then one had better think again.
So much for my critique of backwards induction.
Now what about the omega point ?
The argument takes the quantity theory way beyond Milton Friedman’s wildest dreams. Not only is the long run value of velocity given but so is long run real GDP. The formula is P = M not PY=MV. This matters since some reading this post may not taste of death before real GDP quadruples (kids don’t waste your time — if you only have about 50 years left — you have better things to do with your precious time than read this).
It is possible that massive quantitative easing could cause a lower price level if, in it’s absense GDP declines then stagnates or collapses causing the end of Western Civilization. It is very odd to find the idea that policy can’t affect the long run on Brad’s Blog.
But what about v ? What says that in the long run the interest paid on reserves has to be less than the target Federal Funds rate ? Certainly the Federal Reserve Act doesn’t. The collapse in demand for Fed liabilities depends on that differential which the Fed can keep constant even when the economy returns to normal or over heats.
Huge central bank liabilities and stable prices are inconsistent with the basic principles of fractional reserve banking. What says that reserve requirements must be a fraction ? Again not the Federal Reserve Act. In many other contexts, no actually in exactly this context, Brad discusses the vast powers of financial repression. The Fed can force depository institutions to leave their vast excess reserves in Federal Rserve banks by redefining them as required reserves.
More generally, macroeconomists who try to study the cycle make strong assumptions about the long run, agree on those assumptions, and, sometimes, come to believe those assumptions, exactly because we know so little about the long run. First, there is the idea that the long run is a problem for growth theorists. Second there is little evidence about the long run — there are few non-overlapping long intervals. Third the long run depends on technological progress, which is very hard to understand. So it is treated as exogenous when modelling. So it is assumed to be unaffected by policy. So the stongest possible claims are made about something, because it isn’t understood at all.
In a field where researchers were fundamentally wrong, there would be the most consensus and the strongest confidence exactly where there is the least evidence. In such a (purely hypothetical) field of enquiry, evidence only creates problems.
With inflation targeting, there won’t be an Omega point, because there is 100% hysterisis. Every 1% drop on the price level today causes a 1% drop on the Omega price level.
Inflation-targeting is an open invitation to bubbles in money. It’s like if the government controlled the supply of houses in a country where population is growing, and raised the target price of houses by 1% for every 1% rise in the price of houses today.
Yes, you are right that other stuff might cause bubbles, But we don’t need any of that weirdo stuff to explain the current bubble. Policy induced hysterisis is sufficient.
Adopt an NGDP level-path target now!
Put it another way: an announced NGDP level path target *creates* an Omega point for expectations. That’s how it’s supposed to work.
“Because the long run will come, increases now in the monetary base of sufficient magnitude that are believed* to be permanent will–maybe not now, but soon, and for the rest of our lives, in this long run–produce equal proportional increases in the price level, and thus substantial jumps in the inflation rate as the price level transits from its current to its long-run level.”
I believe that the flaw in that statement is that it assumes that prices can continue to rise in our current situation.
We can not continue to raise employment mostly in the lowest paid wage sectors and expect the economy to return to that of earlier better times. Converting the unemployed to part time waiters and bartenders, polishes the coarse stats but little else. We are spinning down the economy as consumers with well paid jobs lose them, cycle thru a period of unemployment insurance, and are then forced to accept lower paying jobs. That process has slowed but it still continues.
And a quick review of the Total Household Debt over the last 15 years reveals that the changes in debt levels have become much smaller than they were before December 2007. It appears that consumers are maxing out their currently available credit channels. If you lower the consumers’ interest rates to 1% but they fear more debt, or the lenders fear the consumers’ current debt levels or income prospects, then new loans will be minimal.
Consumers can not spend what they do not have and producers will not produce what they can not sell.
If consumers’ wages are stagnant and consumers are carrying close to maximal debt levels, then as prices rise they must reduce discretionary spending. And that would tend to suppress prices.
JimH puts it well and I think he is right. There’s employment and then there is “employment”. Part time waiters and bartenders will not fuel inflation.