« Jobs Figures Support Further Rate Hikes | Main | Here’s a Small Token of Appreciation for Your Many Years of Service »

Friday, August 05, 2005

Filling in the Pieces of the Yield Curve Puzzle

The yield curve conundrum has several popular explanations.  The first is that markets are efficient assimilators of information and collectively low long-term rates are signaling the anticipation of future economic weakness.    The Fed is discounting this explanation.  A second explanation is the global savings glut hypothesis favored by Bernanke.  However, this paper (NBER Subscription req.) by Michael P. Dooley, David Folkerts-Landau, and Peter M. Garber throws cold water on that hypothesis:

Savings Gluts and Interest Rates: The Missing Link to Europe, Michael P. Dooley, David Folkerts-Landau, and Peter M. Garber, NBER WP 11520:  Data for world savings rates do not suggest that an aggregate glut of world savings has depressed US and international interest rates in recent years. Unusual but offsetting changes in savings rates have been limited to three regions: sharp declines in the US have been matched by sharp increases for developing Asia and the Middle East. The world saving rate has increased very little. There are two important features of this change in regional savings behavior. First, three-quarters of the increase in Asian and Middle Eastern savings has been placed in international reserves. Second, all these additional savings have been absorbed by the United States. Even if reserves are mostly placed initially in the US, we would not expect all the savings exported from these high savings regions to remain in the United States. A collapse of expected profits outside the US seems to us a compelling explanation for the US current account deficit and depressed international interest rates.

I haven’t liked the term glut because to me that term implies that the supply and demand for loanable funds are not matched, i.e. there is an excess supply at current rates, not simply that rates have fallen to clear the market.  The Dooley, Folkerts-Landau, and Garber paper contains a third explanation that takes the traditional view that low long-term rates are caused by lack of profitable investment opportunities.  The final hypothesis for the conundrum is the misperception of risk argument that Greenspan has been promoting.  Under this view, which relies upon investor misperception of ARCH effects (periods of high volatility followed by periods of low volatility) and upon biased expectations of risk, investors are accepting risk premia on long-term commitments that are too low.  Assuming irrational expectations to explain the conundrum is not a story I favor.

There are fewer pieces to the puzzle if we can rule out future economic weakness and a global savings glut as explanations, but it still isn’t clear to me exactly how the pieces fit together.  If world savings has not changed, why have rates fallen?  If it isn’t a supply glut, then it must be low demand, i.e. lack of profitable opportunities worldwide.  Is that all it is?  Is it that simple?

Update #1:  There is more on this topic here and by Brad DeLong here, particularly on the excess liquididty versus global saving glut explanations of low long-term rates.  See Angry Bear and William Polley also.]

Update #2:  Here is a conversation from the comments (edited) I hope is instructive:

Mark, ... I have never really understood what is meant by supply and demand for loanable funds. Are we talking stock or flow variables here? Textbooks are rarely very clear on this. They often seem just to say 'demand and supply of loanable funds' and assume the meaning of this is obvious.

If by loanable funds we are talking about asset stocks then what does it mean to say excess supply? Somone has to hold what assets there are. ... Or is the argument that a high level of saving (a flow) means that to keep aggregate demand propped up central banks are pushed to set low short rates ... Which would mean treating 'supply of loanable funds' as a synonym for high savings propensities. ... Would appreciate any response you might offer.

rjw - Income is the flow of water through the bathtub and saving is the stock of water that accumulates (and this can be negative for an individual or for nations, but not for the world as a whole). The faucet is income and the drain is consumption and taxes. What’s left over accumulates in the tub as saving.

Part of that can be held as cash, or as stocks and bonds, as houses, as stamp collections, and so on, then re-liquidated later if needed.

There are two related concepts which are the inverse of each other, one is the demand for financial assets and the other is the supply of loanable funds. That is, those demanding financial assets are the same as those supplying funds to use for loans. The reverse is true too. Those supplying financial assets are doing so to get funds to use, so the supply of financial assets is the same as the demand for funds. There are two models because when you graph the asset supply and demand model the supply and demand curves have the wrong slopes so it's convenient to rename them.

So, when I think of the price of financial assets (and hence the interest rate as they are inversely related), it is with respect to the supply and demand of a stock of financial assets.

The stock will change period to period as savings are added or subtracted from wealth shifting those curves, but at any point in time the supply and demand curves are fixed and the price/interest rate moves to clear the market fairly quickly.

I hope this is helpful.

On your question at the end. If the world saving rate has not fallen, then the argument is that a low level of investment has called a drop off in demand relative to available capacity, leading to central banks setting low rates to stimulate demand? Is that the argument? ...  And then low rates boil down just to accomodative monetary policy on a global scale that has spread over the yield curve (due to anticipated persistent weakness in demand? Is that what you mean by it being dead simple?

rjw - Yes, that is essentially how it works. If the supply of loanable funds is fixed as the article suggests, then weakness in the demand for loanable funds could explain low rates.

Demand depends primarily upon expected future profits for various investment opportunities (with an eye towards other factors such as existing capacity), i.e. the rate of return, relative to the cost of borrowing funds over the period until the enterprise is up and running and the loans can be repaid. So with such low rates the presumption is that the expected rates of return must be even lower (all adjusted for inflation).

It is true that foreign central banks have absorbed a lot of the supply of loanable funds thereby propping up demand. But even then rates have fallen.

mark, thx

Ok - so we are talking demand and supply of financial assets. And thanks for the clarification of terminology. I have a follow up question which I hope you or a commentator might tackle.

It's obvious that at any point in time someone has to hold all assets. In that sense any excess demand can only be for one type of asset over another (I decide I want more bonds and less cash - so I shift my portfolio). And here disequilibria should be short lived as prices should adjust to clear market.

So then low interest rates under this explanation come about simply because there has been some shift in the desire to hold bonds relative to cash - thereby changing the price that clears the bond markets? And people don't want to shift out of bonds into equity maybe, so all the desire for portfolio shift is funneled into demand for bonds (or - to generalize the picture - into assets like housing)

rjw has good questions.  I’ll give this a shot.  To start, we can divide wealth as follows

Wealth = Money + Bonds + Equities + Physical Assets

We often (but not always because whether these assets are substitutes or complements matters) lump all interest bearing assets into the “Bonds” category to simplify things. Then

Wealth = Money + Bonds + Physical Assets

We *used* to say physical assets were constant enough to ignore in the short-run, but if housing is in there we shouldn’t do that anymore, so let’s not. The equation above will serve as our wealth function.

The first cut is to decide how much liquidity (i.e. money, sometimes called speculative balances) to hold. As you note, the interest rate is a key factory here (actually expected future interest rates because capital gains matter). The other cut is to decide how to hold interest bearing assets as bonds (and perhaps bonds versus equities) or houses and other physical assets (for colleagues – I know these aren’t always separable like this, but it’s useful to break it up like this for illustration).

So main point here – the interest rate determines the division of wealth among asset categories for a *given* level of wealth.

But then what is the link here to low central bank rates? What is the conceptual link to those? I feel I'm missing a piece of the puzzle there. I also don't really see how that reasoning ties in with terminology of a savings glut - irt is not as if people want to hold fewer assets - it's just the types of assets they want to hold that has changed.

Central banks change money supplies by increasing or decreasing the supply of financial assets. It may not have changed recently, but that does not mean it wasn’t large already (the existing stock was already large).

As an aside, I suppose I could increase my 'demand' for assets (implicitly) by decreasing my spending out of income to add to my level of assets by saving more. But is that not just another way of saying that the savings rate has shifted? ...  my problem is, if overall demand for assets goes up - people save more - income shifts - so supply and demand curves not stable as desire to hold assets is a function of income levels. That's what is bugging me. Does it make sense to talk about such curves in that context of changing income levels? ...

As income levels change the demand curve shifts up or down. Here’s the reasoning. With more income, there is more saving so wealth increases. With more wealth there is a reallocation over the three categories of wealth (in general all three are expected to rise, but not necessarily equally). So think of this as a shift in the supply of loanable funds (i.e. the demand for financial assets increases).

An increase in the desire to save has the same consequence on wealth, but there is an important difference. Because income has not increased, consumption must fall to compensate for the increase in saving.

    Posted by on Friday, August 5, 2005 at 10:53 AM in China, Economics, International Finance | Permalink  TrackBack (0)  Comments (22)


    TrackBack URL for this entry:

    Listed below are links to weblogs that reference Filling in the Pieces of the Yield Curve Puzzle:


    Feed You can follow this conversation by subscribing to the comment feed for this post.