E-piphany

There was a potentially important development in inflation recently, but one that was generally overlooked.

Perhaps it was mostly overlooked because it is way too early to say that a trend is developed that could cause an adverse inflation occurrence. But I think that the main reason it was overlooked is that monetary velocity is not very well understood. In particular, most people seem to think that money velocity – definitionally, the number of times a unit of money is transacted in a year, on average – is somehow tied to nervousness about the economy. So, when money velocity fell in the global financial crisis, many observers attributed that to savers stuffing dollar bills in the proverbial mattress.

There may be some role for investor/consumer uncertainty in the modeling of velocity, but at best it is a secondary or tertiary cause. The main cause of changes in velocity is simple: when the cost of holding money is high, we work hard to hold less of it, and when the cost of holding money is low, we don’t mind holding more of it. Friedman first noticed this, so it isn’t a new discovery. In monetarist speak, velocity is the “inverse of the demand for real cash balances,” and the demand for such balances depends of course on the relative cost of cash balances relative to other investments.

The chart below shows the power of this relationship. I’m using the 5-year constant maturity treasury rate, but there are obviously other investments that would get thrown into this relationship. But it’s easy to envision the effect here. When interest rates are at 6%, then money does not sit idle for very long or accumulate without limit in your bank account – you will invest those monies in, say, a 5-year CD or Treasury Note, rather than earn basically nothing in a checking account. But when interest rates are at 1%, the urge to do so isn’t as much.

What you can see in the chart is that interest rate moves tend to precede movements in money velocity, which is what we would expect from a causal relationship such as this. So the reason that money velocity plunged in the GFC wasn’t because people were scared; it was mostly because interest rates fell, taking away the incentive to invest longer-term.

Changes in money velocity, of course, tend to cause changes in inflation. MVºPQ, after all, and Q tends to be mostly exogenously determined by aggregate fiscal variables, industrial policy (what’s that?) and the like. Changes in M and changes in V tend to be reflected mainly in P over time.

Also, interest rates are affected by inflation, or more properly by the expectations of inflation. And expectations about inflation tend to follow inflation.

So, the history of the 1980s’ declining inflation can be read like this, without too much of a stretch: declining money growth under Volcker caused declining inflation initially. The decline in inflation tended to cause interest rates to decline. Declining interest rates tended to cause declines in money velocity. Declining money velocity tended to cause declines in inflation…and we were in a virtuous cycle that extended, and extended, and extended, until we were close to zero in interest rates and inflation, and money velocity was as low as it has ever been.

Now, you can see from this chart that interest rates bottomed in 2013, but really have not appreciably risen above the lows, and so money velocity hasn’t reversed its slide although since the beginning of last year the trajectory has been slowing (I suspect some nonlinearity/stickiness of this relationship near zero). But the GDP report from last Friday, combined with recent money growth and increase in the price deflator, implied that money velocity actually rose slightly.

It has nudged higher before, but not by very much. And this is why I am reluctant to make a huge deal about this being the start of something, except that this is the first time since 2008 that there has been a reasonable expectation that interest rates might continue to rise because the central bank wishes it so.

And so I don’t think it’s wrong to consider the “what if” of the next cycle. Normalization of interest rates implies normalization of velocity, and there’s just no way to get appreciably higher velocity without higher inflation. Higher inflation would probably produce higher interest rates, because however much your expectations about inflation are “anchored” they are likely to become unanchored if inflation of 3%-4% starts to print. Higher interest rates could lead to higher velocity, and we have a cocktail for the opposite of the 1980s virtuous interest rate cycle.

This speculation isn’t destiny, and a lot depends on whether interest rates start to move higher and by how much. But there is already starting to be some concern about inflation and the FRBNY’s “Underlying Inflation Gauge” has recently gone to new post-crisis highs (see Chart above, source Bloomberg), so I don’t think it is unreasonable to consider and prepare. Because the best case for the next inflation uptick is that it rises a bit and falls back. But there are elements in place that support a much worse case, and that is a feedback loop through interest rates and velocity. The chances of that outcome are considerably higher than zero.

Note: these articles are now first released on my private Twitter feed, which you can subscribe to for only $10 per month here. Subscribers also get my real-time tweet analysis on the monthly CPI report, which are not on my public feed, and I am working on adding a free chart package to the mix as well.