Well, perhaps we can have one more day of quiet markets. With London closed, volumes were again thin. The most notable movement was the sharp rally in energy markets overnight, led by gasoline. This was provoked by Tropical Storm/Hurricane Isaac, but a huge explosion (killing 41 people) and fire in a Venezuelan refinery helped the rally.

Strangely, though, NYMEX Crude and other non-gasoline products plunged hard once the floor opened. Oil lost about 3% in 45 minutes or so, which is a typical news-reaction signature; however, there seemed to be no news. Bloomberg attributed the selloff to "speculation that Isaac won't do much damage," but I have trouble buying that. Such an explanation doesn't explain the sharpness of the move (it isn't like that realization abruptly swept the entire trading floor and electronic exchanges simultaneously). Moreover, since the storm track as of late last week hadn't been expected to take it far into the Gulf, oil prices hadn't rallied much on the notion that it would do damage. Thus, it doesn't make sense (to me, anyway) that there would be a sharp correction based on a change in opinion that happened over the weekend.

I never did see any news to explain this sudden move, which makes me suspicious. I would expect that if there is any damage at all to Gulf equipment, the President would rapidly tap the Strategic Petroleum Reserve (given any excuse, so that it doesn't appear political), but it doesn't appear likely at this point that there will be any such damage. I probably just missed some news.

Bond rallied, with 10-year Treasury yields down 3.5bps to 1.65%. But 10-year TIPS yields rallied more, 4bps to -0.66%, so that 10-year inflation expectations rose slightly. This seems Bernanke-inspired, since an announcement or hint from the Chairman that QE3 will soon commence would tend to push down nominal yields but push up inflation expectations.

Investors clearly believe that the Chairman will signal QE3 is coming when he speaks on Friday, although the consensus among Bloomberg-polled economists continues to be that it will not happen. I think the fact that markets are discounting such an outcome increases the odds that it will happen, since he probably would prefer not to disappoint markets. Still, he's not the wimp that Greenspan was in this regard. The 'Maestro' very rarely led the music; unlike the typical maestro, he tended to follow and validate what markets were expecting. Bernanke sometimes marches to his own drummer. In this case, though, I think he called the tune, and we're marching to his beat rather than hoping he marches to ours.

Speaking of that beat, the NY Fed blog had a piece today that is interesting for its timing. Entitled "Interest on Excess Reserves and Cash 'Parked' at the Fed," the authors (Gaetano Antinolfi and Todd Keister) sought to argue that:

Because lowering the interest rate paid on reserves wouldn't change the quantity of assets held by the Fed, it must not change the total size of the monetary base either. Moreover, lowering this interest rate to zero (or even slightly below zero) is unlikely to induce banks, firms, or households to start holding large quantities of currency. It follows, therefore, that lowering the interest rate paid on excess reserves will not have any meaningful effect on the quantity of balances banks hold on deposit at the Fed.

While this is true, it is also irrelevant. The question is what would happen to excess reserves compared to required reserves. There are two ways that excess reserves can decline. First, the Fed can reduce the size of the monetary base by selling securities. Second, banks can expand their own lending books, increasing the amount of required reserves. This is, after all, the supposed point of the Fed increasing the size of its balance sheet, a point seemingly lost on these economists. The quantity of balances held by banks at the Fed would be unaffected by a change in IOER, but balances held at the Fed aren't the important variable - M2 (or some other measure of transactional money) is the important variable, and if bank lending increases, then transactional money increases as the multiplier between base money and M2 rebounds.

Here is the path from base money to inflation, in a simplified framework.

• Base money x multiplier = transactional money (M2)
• P = MV/Q

In a flowchart form:

So, here is what has happened since mid-2008

You can confirm for yourself that the change in base money times the change in the multiplier equals the change in M2 by calculating (1 + 2.177)(1 - 0.593)=(1 + 0.293), and that MV=PQ by calculating (1 + 0.293)(1 - 0.163) = (1 + 0.019)(1 + 0.062).

The first thing that jumps out at me here is that the absolute changes are huge. By itself, this should be scary to a policymaker, since unless there's some natural reason that things will unwind the same way they built up, there is a big mess to clean up. And I think the Second Law of Thermodynamics (which says essentially that you can't put the poop back in the goat) makes that optimism ill-placed.

The second thing is that there are two variables in the table above that we are exceptionally poor at forecasting: the money multiplier, and velocity. It seems like the money multiplier responds, or should respond, to incentives to keep money in sterile excess reserves rather than to make loans: Interest on Excess Reserves, the quality of credit in the economy generally, moral suasion, legislation that encourages risk-taking among lenders, and so on. However, we don't have a good idea how these factors interact to affect the multiplier. We do know that between 1994 and 2007, the only time the multiplier moved outside of the range of 8.2 to 8.6 was right at the end of 1999, when the Fed flooded the banking system with reserves 'just in case', but then took them right back out a few months later. In 2007-08, the multiplier rose to 9.4 before plunging once the Fed began QE (see chart below, source Bloomberg).

Velocity, similarly, is something that we don't fully understand the drivers of (although I stumbled on something really interesting just now that I'll share sometime over the next couple of days). We do know that velocity was in the range of 1.6 to 1.8 from 1960 to 1990, before rising to just above 2.1 in the late 1990s and then commencing a long slide to its current level, which is the lowest post-war level on record (see Chart). But we're not totally sure what drives it.

Any way you slice it, the multiplier is at outrageously low levels, as is velocity. It is this combination that has kept the large increase in 'base money' from producing sharply higher inflation. If we had a solid understanding of why these two variables behaved this way in the last couple of years, it may not be as much of a concern - but we don't. And we don't know where they may go next. I would postulate that recent bank lending increases could presage an increase in money velocity, which in any event is probably more likely to move higher into the historical range than continue to move lower. I would suggest that lowering IOER could cause the money multiplier to rise again, while hiking it may cause the opposite (but, admittedly, we're not sure what the marginal impact is of an IOER change, since there has only been one in the course of monetary history in the U.S.).

Let's see why we might care. The table below is derived from the flowchart above, and the assumption that real GDP rises 6% over the next eighteen months (that's not a forecast, but an optimistic assumption that helps to reduce the inflation numbers in the table below. Good growth, I figure, is more likely to be associated with a rebound in velocity, so I want to be fair.). I am sure we all agree that whether or not IOER declines, it is unlikely to rise appreciably in the next 12-18 months during which Bernanke will still be Chairman and the Unemployment Rate will still be over 6.5%-7% at best.

The table calculates the resultant rise in the price level (aggregate, not annualized) if the monetary base is unchanged, GDP rises 6%, and the multiplier and velocity are as shown on the axes. So, for example, if the multiplier rises to 4 and velocity rises to 1.6 (and growth is a healthy 6%, meaning a 4%-6% annualized pace over 12-18 months), then the GDP deflator would rise a gentle 3%, right in the sweet spot of monetary policy, assuming the money base didn't grow further. On the other hand, if the multiplier remains around 3.5 and velocity slips slightly further, to 1.55, then in the absence of money growth we would have a 13% decline in the price level.

More frightening, perhaps, is what would happen if the multiplier rose to 5.5 (which is only 40% of the way back to its 1994-2007 average) and velocity merely returned to the low end of the historical range, at 1.6. Even if the money base didn't grow one penny, prices would have to rise 41%. In fact, I would argue that while this table shows a path to deflation - we need continued weakness in the multiplier and velocity, coupled with a Fed which suddenly puts the brakes on the balance sheet - it shows many more ways to get truly disturbing outcomes on inflation.

Projecting a correction in these metrics over 12-18 months is probably destined to look bad. If we look at a longer time horizon, then GDP growth can help blunt some of the effect of rebounding multipliers and the price increase would also be smeared over a longer time period. But even here, the news is not great. If the economy grows 3% per year for five years (not shown in the table above), and the money base doesn't change, the multiplier returns to 5.5% and velocity gets back to 1.6, as above, then the total price rise of 30% would still represent roughly 6% inflation per year.

The only way you can get tame inflation over an extended period, unless the multiplier and velocity are permanently broken (especially the multiplier), is if the Fed shrinks the balance sheet with the same aggressiveness with which it built it.

I don't see that happening under Bernanke, or under most potential replacements for him.