Tuesday, July 26, 2011

Some good old fashioned Keynesianism and Default

If you were on the D.C. metro last night and you saw me tracing out IS-LM curves in the air with my fingers with a puzzled look on my face, it was because I was thinking about this post. Nick Rowe asks what a simple IS-LM model would imply about the consequences of default. He says that the IS curve would shift to the right, the IS-LM equilibrium would shift to the right, and the IS-LM model would say we would be better off.

This is odd, right? In the guts of the IS curve, ignoring what's going on with the LM curve, the equilibrium shifts left and the equilibrium savings/investment level decreases. The equilibrium interest rate is higher but because of a negative shock to the supply of savings. Investment demand equilibrium decreases and there's no clear mechanism for an increase in government or consumption spending, so the ultimate salutary result is puzzling.

Most of the responses have focused on the LM curve. Brad DeLong says it should shift money demand out as well, canceling the shift in the IS curve. This isn't implausible to me. Money demand is a security blanket, so the idea that it would increase makes sense, particularly given Paul Krugman's point that the primary difference between inflation and default (which Nick considers comparable) is that default doesn't directly affect the value of cash the way inflation does.

My reaction was somewhat different. Treasuries are assets, and if those assets suddenly become worth much less it's going to impact demand. How is this any different from mortgage backed securities exploding in people's faces (oh right... it's different because this isn't inevitable - but there's nothing analytically different about it)? So my first thought wasn't Brad DeLong's thought that adjustments in the demand for money would cancel out any counter-intuitive expansion due to a higher risk premium. My thought was that within the IS curve itself decreasing demand would be at least as strong as increasing interest rates due to a supply of loanable funds curve that shifts left.

Nick says "this is like an increase in inflation", but it's really not is it? Inflation has some nice side effects for debtors. Inflation can have the consequences of spurring investment. Does default do that? Put it this way - if we can just think of inflation and default as comparable, why do so many governments seem to prefer inflation to default? If they were exactly the same wouldn't they default more often?

Does this make sense at all?

UPDATE: Hmmmm... thinking more about it and I'm not really sure it's all that different from inflation after all. Anyway - thoughts welcome.


  1. Daniel,

    This is a bit off topic, but I would like to ask you what you think of that interesting chart I found on this site: http://www.wellenreiter-invest.de/WellenreiterWoche/Wellenreiter110723.htm

    Unfortunately the article is in German. Therefore I want to summarize the basic point of it:
    The graph shows that the US-CPI (red line), commodity index (blue line) and the interest on 10 year US-Bonds (green line) highly correlated in the past until 1981. Then suddenly the correlation completely broke up. Yet at least the commodities returned to correlation with CPI again with a strong move in the last decade. But the interest on 10Y Bonds is still “out of step” as argued in the article. They of course argue that this is unnatural and also the green line will sooner or later have to return into correlation mode again, which could be triggered by a stuttering economy of China.

    I really would like to have your take on it.
    - What is the reason that the correlation stopped in 1981? Maybe the reason that in 1981 the world accepted the new pure fiat money regime?
    - Are there real factors that will bring the interest on 10Y Bonds back in line, or is this correlation over now once and for all?

    Maybe this chart is not interesting at all for you, then sorry for inconvenience.

  2. 1. I used to read German semi-sort-of-OK, but I'm lost now.

    2. Off the top of my head, it seems like we ought to be looking at change in the CPI rather than the CPI, right? CPI starts growing with the creation of the Fed (understandably), and there's a presumption of price growth. If you look at inflation it tracks interest rates very closely through the 20th century.

    I'm not sure what the logic of looking at CPI rather than inflation is.

    re: "Maybe the reason that in 1981 the world accepted the new pure fiat money regime?"

    In the way that you are meaning this, this would have been Nixon, not Reagan - right?

  3. 1: Kuehn looks like a former German name. German ancestors?

    2: I am not sure why it would be better to look at the change in CPI rather than CPI. I mean yes inflation was low and quite stable long time after 1981, which is why the CPI is rising quite linear and less steep than before. But the point of the article is that a CPI is highly correlated with commodities, and commodities correlate with interest rates on 10Y Bonds. At least they did before. Yet although commodities have risen heavily since the last decade interest on Bonds should have risen too. But they have decreased further contrary to their behavior of hundreds of years before. Isn’t that strange? Now the question is: Is this without consequence because of maybe the current design of monetary system, or is there pressure building up from real factors?

    “In the way that you are meaning this, this would have been Nixon, not Reagan - right?”

    No. Although Nixon “closed” the Gold window, it was 10 years later that the correlation broke. Exactly when the precious metals bull market ended. I consider Volcker as the man who brought trust back into the USD.

  4. You also have to consider all the nasty intertemporal coordination problems you get with inflation. :)


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