Scott Freelander recently asked me about a Bernanke post, which appears to be guilty of reasoning from a price change:

The U.K. economic slowdown to come will be exacerbated by falling asset values (houses, commercial real estate, stocks) and damaged confidence on the part of households and businesses. Ironically, the sharp decline in the value of the pound may be a bit of a buffer here as, all else equal, it will make British exports more competitive.

Here I’d probably cut Bernanke a bit more slack than Scott, as the phrase “all else equals” seems a nod in the direction of the dangers of reasoning from a price change.  The pound fell sharply when the Brexit vote was announced, because of an anticipated decline in the demand for pounds.  Brexit will reduce the foreign demand for British goods, services and assets.  Since one needs pounds to buy British stuff, this reduces the value of the pound, as well as the quantity of exports. Think of it as a leftward shift in the demand for pounds, on an S&D diagram. Bernanke presumably meant that British exports fall by less than if the BoE had pegged the pound, while demand was shifting left.  That is correct.

One other point.  I recall one recent example where the pound fell a couple of cents on expansionary talk from Mark Carney.  That can be viewed as a positive shift in the supply of pounds, which would indeed boost exports.

Here’s another example that Scott noticed, from the same post:

In the United States, the economic recovery is unlikely to be derailed by the market turmoil, so long as conditions in financial markets don’t get significantly worse: The strengthening of the dollar and the declines in U.S. equities are relatively moderate so far. Moreover, the decline in longer-term U.S. interest rates (including mortgage rates) partially offsets the tightening effects of the dollar and stocks on financial conditions. However, clearly the Fed and other U.S. policymakers will remain cautious until the effects of the British vote are better sorted out.

Long-term rates probably fell due to a decline in expected NGDP growth after Brexit (or maybe a greater preference for safe assets).  Presumably Bernanke meant that the drop in long-term interest rates would be more expansionary than if the Fed had pegged those rates by selling T-bonds, right as expected NGDP growth in the US was declining.  Again I’m cutting Bernanke some slack, as he’s obviously a brilliant economist and in his memoir I recall him saying that rising long-term interest rates during QE could actually be a sign that it was working.  So I think his views are not far apart from mine.

Nonetheless, I’m pretty fanatic on the “never reason from a price change issue”, and I feel that even while Bernanke is aware of all the points I just made, talking about the effect of lower interest rates and lower exchange rates can tend to mislead the public.  In another recent post I said:

I certainly agree with the 38 out of 40 economists who view anti-trade deficit arguments as reflecting ignorance of the most basic ideas in EC101. And yet, according to the Council on Foreign Relations, guess who else is ignorant of EC101?

Since April, Treasury has been applying a quantitative framework to determine if a country is managing its currency inappropriately for competitive advantage–that is, keeping it undervalued. Japan already meets two of the three criteria–a bilateral trade surplus with the U.S. over $20 billion, and a current account surplus greater than 3 percent of GDP–and will meet the third if intervention exceeds ¥10 trillion in a twelve-month period. This is not a high threshold historically–Japan sold ¥14 trillion in 2011 and ¥35 trillion in 2003-4.

So apparently those highly educated bureaucrats at the Treasury, with their 6 figure incomes and their posh DC lifestyles, are actually a part of the ignorant masses that are pushing Trump-style populism. And in fact they are pushing nonsense, the “quantitative framework” has no more support in economic theory than astrology has in high-level physics. So if the public has been reading articles for decades and decades about how our Treasury officials valiantly try to protect us from evil Asian exporters, is it any wonder that the now are susceptible to the arguments of right and left wing populists?

I worry when experts talk about the expansionary impact of a lower exchange rate, or a lower interest rate.  This recent Forbes piece is an example of what may result:

After Friday’s market close, people remarked that both the bond market and the stock market were at all time highs.

It’s not supposed to work that way.  Now, it is a common misconception that bonds always are negatively correlated with stocks.  Actually, over the long term, they have a correlation of zero with stocks.  But they spend most of their time in one of two regimes, either strongly positively correlated or strongly negatively correlated.  Over time it works out to be zero.  Yet here we are, with stocks and bonds on the highs.

David Zervos, market strategist at Jefferies, commented that “Central banks may finally be taking this too far.”  I think central banks started taking things too far in 1913, but yes, with nearly every financial asset in the stratosphere, you could easily come to the conclusion that there has been too much monetary easing.  I am not the first to say that central banks are addicted to higher asset prices.  It’s hard to imagine a scenario where they willingly let the markets deflate.

We’ve been having a lot of bubbles in recent years (a feature of a world populated with central banks), from the dot-com bubble in 2000 to the housing bubble in 2007 to what people are calling the “central bank bubble” or “the everything bubble” now.  Chances are, this could be the biggest bubble of all, and perhaps the most dangerous.

A few years ago, I predicted that in the future there would be almost non-stop complaints about bubbles.  People would see them everywhere.  That’s because low interest rates are the new normal, and thus P/E ratios, price to rent ratios, etc., will be higher than in the past.  It will look like there are bubbles everywhere, but of course bubbles don’t actually exist.

Part of the problem is that the public thinks it’s been told that low rates are easy money, which should boost asset prices.  So they see this as a central bank phenomenon, even though the lowest rates are in places (like Switzerland) where money has been tightest, and the higher rates are in easier money places like Australia.  The public misreads posts like the Bernanke example I just cited, and learns the wrong lesson.  That why I want economists to stop talking about the causal impact of a change in interest rates, inflation or exchange rates, and start talking in terms of the causal impact of changes in NGDP growth, where expected NGDP growth represents the stance of monetary policy.

Let me remind you of some earlier words of wisdom from Bernanke:

The imperfect reliability of money growth as an indicator of monetary policy is unfortunate, because we don’t really have anything satisfactory to replace it. As emphasized by Friedman (in his eleventh proposition) and by Allan Meltzer, nominal interest rates are not good indicators of the stance of policy, as a high nominal interest rate can indicate either monetary tightness or ease, depending on the state of inflation expectations. Indeed, confusing low nominal interest rates with monetary ease was the source of major problems in the 1930s, and it has perhaps been a problem in Japan in recent years as well. The real short-term interest rate, another candidate measure of policy stance, is also imperfect, because it mixes monetary and real influences, such as the rate of productivity growth. In addition, the value of specific policy indicators can be affected by the nature of the operating regime employed by the central bank, as shown for example in empirical work of mine with Ilian Mihov.

The absence of a clear and straightforward measure of monetary ease or tightness is a major problem in practice. How can we know, for example, whether policy is “neutral” or excessively “activist”? I will return to this issue shortly. . . .

Do contemporary monetary policymakers provide the nominal stability recommended by Friedman? The answer to this question is not entirely straightforward. As I discussed earlier, for reasons of financial innovation and institutional change, the rate of money growth does not seem to be an adequate measure of the stance of monetary policy, and hence a stable monetary background for the economy cannot necessarily be identified with stable money growth. Nor are there other instruments of monetary policy whose behavior can be used unambiguously to judge this issue, as I have already noted. In particular, the fact that the Federal Reserve and other central banks actively manipulate their instrument interest rates is not necessarily inconsistent with their providing a stable monetary background, as that manipulation might be necessary to offset shocks that would otherwise endanger nominal stability.

Ultimately, it appears, one can check to see if an economy has a stable monetary background only by looking at macroeconomic indicators such as nominal GDP growth and inflation. On this criterion it appears that modern central bankers have taken Milton Friedman’s advice to heart.  [emphasis added]

That last sentence seemed true in 2003, but obviously not today.  What happened?

See if this sounds familiar. Halfway through a long expansion, industrial production levels off, after years of rapid growth. This is blamed on two factors, declining oil prices and a strong dollar.  I could be describing the past year, or I could be describing 1985-86:

Notice how industrial production leveled off for a couple years, before resuming its rise:

And here is the oil price, which fell gradually in the 1982-85 period and then plunged in 1985-86:

And here is the trade-weighted dollar, which soared in value, and then plunged:

My hunch is that the dollar was more important than oil.  Tight money led to a slowdown in NGDP growth and a strong dollar.  NGDP growth (year over year) slowed from 12.4% in 1984 Q1, to only 4.9% 1986 Q4. Eventually the Fed eased policy, and NGDP accelerated.  The 1980s boom resumed.  It actually eased a bit too much in the late 1980s, which set up the 1990-91 recession.

Does this mean that history will repeat? No. But it’s an interesting comparison.

PS.  While writing this I forget that Caroline Baum made a similar observation in her book “Just What I Said“, where she pointed out:

This isn’t the first time that a change in oil prices has been regarded as a tax increase or tax cut and anointed with the ability to help or hinder economic growth. Time-trip back to early 1986, when oil prices plunged to $10 a barrel in April from $30 at the end of 1985. This was hailed as good news – a tax cut! – for consumers, which was guaranteed to boost US economic growth.

It didn’t turn out that way.  Following the plunge in oil prices, gross domestic product growth slipped to an anemic 1.7% in the second quarter of 1986 . . .

Ah, recall the days when 1.7% RGDP growth was anemic, not above trend.

This is an issue that comes up all the time; so let me try to put things in perspective:

1.  Asset prices are procyclical, dropping sharply in slumps like 1929-33, 1937-38, 2000-02, and 2007-09.  (This entire post focuses on real asset prices.)

2.  Interest rates are also strongly procyclical.  Thus interest rates and asset prices often tend to move in the same direction.

3.  When interest rates fall for reasons other than the business cycle, asset prices tend to rise.  Thus if rates fall due to a global savings glut, it will tend to raise asset prices.  That’s the most likely explanation for the asset price boom of 2009-15.

4.  What about monetary policy?  Asset prices tend to rise when monetary policy is easier that expected.  But on closer examination it seems like real stock prices don’t “like” either easy or tight money, stocks like stable money.  Stocks do very well in low inflation booms (1920s, 1980s, 1990s, etc.) and do poorly during either deflation (early 1930s, 1938, 2009) 0r high inflation (1966-81).  Thus it would be more accurate to say that stable money is good for stocks, not easy money.  The January stock slump is accompanied by lower interest rates, but not caused by lower interest rates.  It’s caused by the thing causing lower interest rates (lower NGDP growth expectations—and hence tighter money.)

5.  David Glasner did a study that suggests stocks tend to rise strongly on easier money (bigger TIPS spreads) precisely when deficient AD is a serious problem.  That shows that asset markets “root for” sound monetary policy.   In the 1980s, asset prices rallied on tighter money (lower NGDP growth.)

Conclusion:  Monetary policymakers are probably unable to create bubbles, even if they try.  That’s because asset prices will be highest when policy is boring and appropriate.  Monetary policymakers can create asset price crashes by doing crazy stupid things, but they cannot push real asset prices higher than they would be with sound policy.  Trump might say that asset prices like situations where “America wins.”

This post is similar to a post I did over at Econlog (which is the better post, BTW.)  There I pointed out that wage growth usually slows when RGDP growth slows, but paradoxically lower wage growth causes higher RGDP growth.  And interest rates usually fall when NGDP growth slows, but paradoxically when the Fed cuts its target rate that causes NGDP growth to rise.  Now we can see that interest rates usually fall when asset prices crash, but paradoxically a Fed target rate cut usually boosts asset prices.

The failure of reporters to internalize the implications of these paradoxes explains much of the nonsense you read in the media, such as the current popular theory that Chinese devaluation is deflationary.  The real issue is, “does Chinese devaluation cause other central banks, in places like Brazil, to run more deflationary monetary policies?”  Or, “does Chinese devaluation reflect weak growth, cause stronger growth, or both?  And what is the independent effect of each of those factors?

PS.  A comment by Kevin Erdmann over at Econlog anticipated some of my thinking on this issue.

I will be traveling today, not much time for comments.

Most people seemed to think my previous post was crazy, and looked for weaknesses.  A few perceptive observers, such as Nick Rowe and Jonathan, noticed that it had the same implication as the IS-LM model.  Some people wrongly assumed I was simply talking about correlation, whereas I was claiming that lower interest rates cause falling NGDP.  Some wondered why that is not reasoning from a price change.

I think the best way to address this confusion is to start with the classic 1988 paper by Barsky and Summers.  They claim that the “Gibson Paradox” is caused by the fact that low interest rates are deflationary under the gold standard, and that causation runs from falling interest rates to deflation.  Note that there was no NGDP data for this period, so they use the price level rather than NGDP as their nominal indicator.  But their basic argument is identical to mine.

The Gibson Paradox referred to the tendency of prices and interest rates to be highly correlated under the gold standard. Initially some people thought this was due to the Fisher effect, but it turns out that prices were roughly a random walk under the gold standard, and hence the expected rate of inflation was close to zero.  So the actual correlation was between prices and both real and nominal interest rates.  Nonetheless, the nominal interest rate is the key causal variable in their model, even though changes in that variable are mostly due to changes in the real interest rate.

Since gold is a durable good with a fixed price, the nominal interest rate is the opportunity cost of holding that good.  A lower nominal rate tends to increase the demand for gold, for both monetary and non-monetary purposes.  And an increased demand for gold is deflationary (and also reduces NGDP.)

Of course that’s just the demand for gold, what about the supply?  It so happens that the supply of gold was fairly stable under the gold standard, rising by about 2% per year, whereas the demand for gold was much more unstable.  Thus changes in the value of gold (which was the inverse of the price level under the gold standard) were mostly caused by shifts in the demand for gold, which were in turn caused by changes in nominal (and real) interest rates.

An interesting question is how these changes impacted the real economy.  I would argue that sticky wages caused the fall in NGDP to result in a fall in hours worked.  A real business cycle proponent might deny that, and claim that whatever caused real interest rates to decline, also caused workers to want to take long vacations.  But anyone who denies the RBC model, and believes wages are sticky, should agree with me; causation goes from interest rates to NGDP to hours worked, even if the initial change in real interest rates was caused by a real shock.  Falling NGDP has an independent effect on hours worked even if caused by a real shock, just as a gunshot wound hurts someone who already has pneumonia.

As far as the claim that this is just IS-LM, I suppose that’s true, but it didn’t stop Barksy and Summers from getting their paper published in the JPE, nor did it prevent Tyler Cowen from calling it an enjoyable paper that addressed an interesting “puzzle”.

The puzzle of why the economy does poorly when interest rates fall (such as during 2007-09) is in principle just as interesting as the one Barsky and Summers looked at.  Just as gold was the medium of account during the gold standard, base money is currently the medium of account.  And just as causation went from falling interest rates to higher demand for gold to deflation under the gold standard, causation went from falling interest rates to higher demand for base money to recession in 2007-08.

There’s no “trick” in my previous post, I meant what I said.  But I’m not surprised that people are confused; after all, didn’t most economists believe the Fed was pursuing an “expansionary” policy in 2008?  Funny how those “expansionary” policies are almost always associated with recessions.  People tend to wrongly equate interest rate movements and “monetary policy”.  Most changes in interest rates reflect changes in the macroeconomy (growth and inflation) not monetary policy. When rates fall, the Wicksellian rate is usually falling faster, which means money is getting tighter in the NK model.

And finally, a word on reasoning from a price change.  Suppose you claimed that low rates should lead to more housing construction, or more investment in general.  That would be reasoning from a price change.  You’d be talking about the impact of the change in a price, on the quantity in the very same (credit) market.  Obviously if low rates are caused by more supply of saving, then the quantity of investment will rise, and if caused by less demand for investment, then the quantity of investment will fall.  But here’s what I’d like to emphasize.  Lower rates will reduce velocity and NGDP regardless of whether they are caused by more supply of saving or less demand for investment.  And that’s because interest rates are not the price of money, they are the price of credit.  So interest rates become a shift variable in the money market.  Lower rates shift the demand for money to the right, which raises the value of money, which is deflationary.  So there’s no reasoning from a price change in that case.  Of course if I didn’t hold the supply of base money fixed, it would be reasoning from a price change.

No, this is not a NeoFisherian post.  I’m not claiming that a Fed policy that depresses interest rates is contractionary, I’m claiming lower interest rates are contractionary, ceteris paribus.  And ceteris paribus in this case means for any given money stock.  For simplicity, we’ll start with a simple model–no IOR— and then bring in IOR later.  And in doing so I’ll answer a question a commenter asked me: Is talking about the effect of IOR an example of reasoning from a price change?

Let’s start with this identity:

M*V = P*Y

Where M is the base and V is base velocity.  Now let’s build a model:

M*V(i) = P*Y

Since V is positively related to i, lower interest rates are contractionary, they reduce V and hence NGDP, AKA aggregate demand.  Larry Summers once wrote an article (with Robert Barsky) pointing this out, but only for the gold standard period.

So why do people not named Summers and Sumner not know this?  There are several reasons:

Sometimes, not always, reductions in interest rates are caused by an increase in the monetary base.  (This was not the case in late 2007 and early 2008, but it is the case on some occasions.)  When there is an expansionary monetary policy, specifically an exogenous increase in M, then when interest rates fall, V tends to fall by less than M rises.  So the policy as a whole causes NGDP to rise, even as the specific impact of lower interest rates is to cause NGDP to fall.

2.  Another problem is the Keynesian model, which hopelessly confuses the transmission mechanism.  Any Keynesian model with currency that says low interest rates are expansionary is flat out wrong.  That’s probably why economists were so confused by 2008.  Many people confuse aggregate demand with consumption.  Thus they think low rates encourage people to “spend” and that this somehow boosts AD and NGDP.  But it doesn’t, at least not in the way they assume.  If by “spend” you mean higher velocity, then yes, spending more boosts NGDP.  But we’ve already seen that lower interest rates don’t boost velocity, rather they lower velocity.

Even worse, some assume that “spending” is the same as consumption, hence if low rates encourage people to save less and consume more, then AD will rise.  This is reasoning from a price change on steroids!  When you don’t spend you save, and saving goes into investment, which is also part of GDP.  Now here’s were amateur Keynesians get hopelessly confused.  They recall reading something about the paradox of thrift, about planned vs. actual saving, about the fact that an attempt to save more might depress NGDP, and that in the end people may fail to save more, and instead NGDP will fall.  This is possible, but even if true it has no bearing on my claim that low rates are contractionary.

To see the problem with this analysis, consider the Keynesian explanations for increases in AD.  One theory is that animal spirits propel businesses to invest more.  Another is that consumer optimism propels consumers to spend more.  Another is that fiscal policy becomes more expansionary, boosting the budget deficit.  What do all three of these shocks have in common?  In all three cases the shock leads to higher interest rates.  (Use the S&I diagram to show this.)  Yes, in all three cases the higher interest rates boost velocity, and hence ceteris paribus (i.e. fixed monetary base) the higher V leads to more NGDP.  But that’s not an example of low rates boosting AD, it’s an example of some factor boosting AD, and also raising interest rates.

Again, I defy you to explain how low rates can boost NGDP, ceteris paribus.  If you think you have an explanation, it’s probably something that confuses consumption with total spending on NGDP.  An explanation that wrongly assumes the public’s desire to spend more on consumption, as a result of lower interest rates, is expansionary for NGDP.  Yes, an exogenous change in M will often cause short term rates to move in the opposite direction, but how often do you see exogenous changes in M?  If we were operating in a normal economy, with say 3% or 4% interest rates, and I told you rates would fall to zero in the next 12 months, would you predict a recession or boom?  Obviously a recession.  Yes, if the Fed perversely cut rates to zero in an otherwise stable economy, via fast growth in the base, that would be expansionary.  But more than 100% of the expansionary impact would come from the rise in the base (hot potato effect), and less than zero from the lower rates.

There is simply no mechanism in macroeconomics where low rates actually CAUSE more NGDP.  None.  Nada.  Lower rates reduce velocity, and that’s contractionary.  It’s not about “spending”, unless by “spending” you mean velocity.  If you mean “spending” vs. saving then you are hopelessly off track.  You aren’t even in the right train station.

Now for the hard part.  The Fed recently raised the fed funds rate, and did so without lowering the base.  So am I claiming that the Fed’s decision was expansionary?  No, but I wouldn’t blame you for seeing a contradiction here, especially if your last name is “Murphy.”

Suppose the Fed had raised the fed funds target, without raising IOR.  What then?  Then they would have had to reduce the monetary base enough to make the rate increase stick.  How much?  Fasten your seatbelts—by almost $3 trillion.  That’s right, without IOR, to even get a measly quarter point increase, they would have had to withdraw almost the entire previous QE (except the part that went into currency held by the public.)

Instead the Fed did something else, they raised IOR.  Even though IOR includes the term ‘interest,’ as part of its acronym, it actually has absolutely nothing to do with market interest rates as I’ve been discussing them so far.  It’s better to think of IOR as a tax/subsidy scheme.

Previously I claimed that higher interest rates are expansionary.  They are.  But higher IOR really is contractionary.  That’s why the New Keynesians love IOR so much.  It helps make their false “non-monetary” models of the economy less false, indeed sort of truish.  It really is true that higher IOR is contractionary.  So why the difference?  Let’s return to the simple model above.  I said that model applied to a world of no IOR.  If IOR exists, then the more general model is:

M*V(i – IOR) = P*Y

That is, velocity is positively related to the difference between the market interest rate and the interest rate on money.  This gap is the opportunity cost of holding reserves.  I wish there were no IOR, if only because it would make monetary analysis so much simpler.

In order to make monetary policy more contractionary, the Fed merely needs to shrink the gap between i and IOR.  That reduces the opportunity cost of holding base money, which causes more demand for base money (as a share of NGDP), which is contractionary.  Consider the weird situation we are in:

1.  Almost everyone assumed higher rates were contractionary, but almost everyone was wrong.

2.  Now IOR comes along, and higher IOR really is contractionary.

Now I fear that it will be even harder to slay the Keynesian dragon; the model now seems superficially even more plausible. If I was a conspiracy nut I’d think it was all a plot to make Woodford’s moneyless models appear more accurate.

To summarize, IOR is not really a market price, it’s a subsidy on base money, and negative IOR is a tax on base money. Just as it’s OK to reason from an increase in excise taxes on gasoline, it’s OK to reason from a change in IOR.  (Of course you also need to consider other changes that are occurring in monetary policy, as is always the case.)

So if lower interest rates are not the reason that monetary stimulus is expansionary, then what is the reason?  Why do more people want to go out and assume car loans when the Fed cuts interest rates via an easy money policy?  Here’s why:

1.  If the Fed lowers rates via an increase in the base, then more base money raises NGDP via the hot potato effect (AKA, laws of supply and demand).  Interest rates play no role, indeed NGDP would rise by even more if rates (and velocity) didn’t fall.

2.  The higher NGDP causes more hours to be worked, due to sticky nominal hourly wages.

3.  More hours worked means more output and more real income.

4.  Say’s Law says supply creates its own demand.  So as workers and capitalists produce more output and earn more income, they go to car dealers to splurge with their sudden newfound wealth.  Interest rates got nuthin to do with it.  Saving (and investment) as a share of GDP actually rises during booms created by monetary stimulus.  It’s not about “spending” (as in consumption), it’s about actual spending on C+I+G+NX, i.e. NGDP.

5.  Of course all this happens simultaneously, as we live in a Ratex world.

PS.  I’m begging you, don’t try to explain what you think is wrong with Say’s Law, unless you want me to be as insulting as possible in response.