Saturday, December 4, 2010

Lee strikes again

Lee has a new post on Modus Tollens on quantitative easing that makes a point I had never thought of before and that I don't think I've read anywhere else (he's probably made it a million times and it's just finally clicking for me).

I've noted elsewhere that I see three arguments for QE2 - all good ones, I might add. It can:

1. Lower long-term rates to boost investment demand (classic Keynesian)
2. Boost nominal expenditures (classic monetarist, widely adopted Keynesian)
3. Increase inflation to deal with wage and price rigidties (New Keynesian)

One thing I haven't comprehended, though, is how simply increasing nominal income guarantees any substantial impact on employment. To put it bluntly "we're in a liquidity trap" (I think we probably are), but to put it more accurately and less controversially, we simply have very high liquidity preference - which will do the trick of getting you below full employment even if you're not in a liquidity trap. What does handing people money really do when liquidity preference is a problem? It might help, but there's no guarantee even a substantial portion of it is going to be put to work.

The answer is quite obvious in retrospect. Longer term bonds don't simply have higher rates that have more flexibility to be pressed down. Longer term bonds don't simply play a role in long-term inflation expectations. The other thing about longer-term bonds that gives a boost to nominal expenditures traction is that long-term bonds aren't all that great for satisfying liquidity preference - they are not as close substitutes for money as short-term bonds. Lee writes:

"Short term bonds with near zero interest rates are extremely close substitutes for money, and so purchasing short term bonds may increase the money supply, but is also likely to increase money demand in proportion, and so any excess demand for money will remain unchanged. Purchasing longer term bonds with higher interests rates, however, exchanges quite different assets. Households and firms that sell long term bonds are unlikely to hold their new money, but instead they will begin spending it on various consumer and capital goods. (Initially, they likely will be reluctant to lend for the same reasons as banks)."

OK - one more (probably dumb) question. Long-term bonds are not like cash in the way short-term bonds are (and even short-term bonds, as Scott Sumner regularly points out, aren't perfect substitutes). But long-term bonds do afford a liquid market. If the problem is a demand for liquidity - an increase in liquidity preference - might it be plausible that people who sell their long-term bonds were still holding them for liquidity purposes, and therefore would still be unlikely to go out and spend? Or am I still misunderstanding something.

Anyway - I've inched closer to the "eh - fiscal policy seems to be more trouble than it's worth" position. Not there yet. But an inch closer.

8 comments:

  1. Daniel,

    It's a matter of degree. Suppose that people who sell short-term bonds hold 90% of the money instead, but people who sell long-term bonds hold only 50 percent. Given the same amount of monetary expansion by the Fed, purchases of longer term bonds will give more bang for the buck, so to speak.

    Following this logic, one might ask why the Fed does not instead buy private junk bonds, with very high interest rates. Presumably they a higher level of total nominal income for every dollar for every dollar created. This is probably true, but it would also expose the Fed to significant risks and raise other difficult political questions.

    Btw, thanks!

    ReplyDelete
  2. Sorry for the garbled nature of the above comment. I have been awake since 4:30am (opening a gas station, would you believe?). Here is how the comment should have been:

    It's a matter of degree. Suppose that people who sell short-term bonds hold 90 percent of the money instead, but people who sell long-term bonds hold only 50 percent. Given the same amount of monetary expansion by the Fed, purchases of longer term bonds will give more bang for the buck, so to speak.

    Following this logic, one might ask why the Fed does not instead buy private junk bonds, because they are very poor substitutes for money. Presumably, they offer a higher level of total nominal income for every the Fed creates. This is probably true, but it would also expose the Fed to significant risks and raise other difficult political questions.

    ReplyDelete
  3. Gah, that's supposed to read: "for every dollar the Fed creates."

    ReplyDelete
  4. Lee -
    No - the first comment was entirely clear and exactly what I was thinking. I'm certainly not saying long-term bonds present the same liquidity advantages as short-term bonds. But long-term government bonds especially are held by those with higher liquidity preference than holders of other long-term assets.

    And clearly for that reason QE2 is a good idea. But the question of the best policy is clearly something of a chicken-vs-the-egg question. You think the lack of aggregate demand is a symptom not a cause. I think the money demand is a symptom not a cause. In reality we're probably both right and there's some sort of simultaneous determination here.

    I'm not quite sure how to arbitrate that, except to reason through it. What would cause an increase in preference for liquidity? I would say, a negative shock to the perceived value of investment opportunities, which would suggest that the money demand is symptomatic of a fall in aggregate demand and not vice versa. I'm showing my Keynesian colors here, but I just don't see the point of wanting money for is own sake. Money itself is pointless - it has value for what you can do with it. You want more of it because you intend to buy more (THAT'S clearly not the case now), you want more because you want to speculate on its value (maybe some of this with deflation or unexpected disinflation expectations), or you just hold it because you don't want anything else that you'd exchange it for. We just went through a crisis where a variety of assets were revealed to be worth much less than originally thought, which (1.) raises questions in peoples' minds about other assets, and (2.) reduces wealth which reduces demand. This alone is an aggregate demand shock, but it also makes potential investors feel like there aren't many investment opportunities out there. So what the hell - sit on your money. You've got nothing better to do with it.

    So to me, the "money demand is a symptom not a cause" makes a lot of sense.

    What is the story to suggest money demand is not just a symptom?

    You've already given me one "aha!" moment this morning that marginally shifted my Keynesian/monetarist ratio a little closer towards monetarism. Care to do it again? :)

    ReplyDelete
  5. Daniel,

    I do not think a decline in the perceived value (or increase in perceived risk) of investment opportunities should reduce aggregate nominal demand. The peculiar heart of my position is that aggregate nominal demand ought not to change much -- it should be insensitive to changes in real supplies and demands. The economy could be going through a tumultuous real supply shock, or an radical shift in relative demands, or a fall in aggregate real demand, and I would still recommend the Fed keep aggregate nominal demand relatively constant (or growing at a constant rate).

    This may seem remarkably simplistic (and I am oversimplifying a little), but it was never my assumption. Instead, it is the result of following through the logic of different possible scenarios. I am still trying to figure out a counter-example.

    ReplyDelete
  6. Perhaps I missed this in previous points, but what is your evidence for: "To put it bluntly "we're in a liquidity trap" (I think we probably are), but to put it more accurately and less controversially, we simply have very high liquidity preference."

    ReplyDelete
  7. As a quick response, JP Koning, the position of the fed funds rate, the fact that prices seem pretty insensitive to the monetary expansion, and the drop in investment despite the clear availability of loanable funds strike me as pretty indicative of a liquidity trap, whether you take the horizontal LM curve interpretation or the inelastic loanable funds market curve interpretations.

    That having been said:

    1. There are probably others out there that could justify the claim better, and

    2. A "trap" is a tricky thing to pin down - I'd prefer to just say we have a large shock to liquidity preference. Also when we say "liquidity trap" some people have the tendancy to respond "but you can still do monetary policy so there's not a liquidity trap", which to me is a non-sequitor. I am not of the view that you can't do monetary policy in a liquidity trap, although it certainly weakens it.

    ReplyDelete
  8. Ok, the fed funds rate has been and continues to be low. I see your point. Why are you so sure that this represents a desire for liquidity? Maybe it represents a desire for safety of capital - after all, reserves are the liabilities of one of the most solid institutions in the world.

    If people want liquidity, they probably want US dollars, the most liquid instrument in the world. Yet the US dollar is far cheaper than it was in 2008 relative to the yen, euro, c-dollar etc. Hardly a sign of increased demand for liquidity.

    Lastly, the S&P just hit a new high and is up about 75% since early 2009. This shows that at least some people have been moving out of liquid dollars into less liquid stocks.

    Anyways, I'm neither here nor there on this issue. I just think that determining people's level of desire for "liquidity" is very difficult. If anything, liquidity should be conceptualized relativistically, ie. more or less liquidity is desired than before, and not absolutely, ie. we are or are not in a liquidity trap.

    ReplyDelete

All anonymous comments will be deleted. Consistent pseudonyms are fine.