Interest Rate Risk in Credit Markets

Recent events have stimulated interest in the joint behavior of prices and quantities in credit markets. Data sources such as the Federal Reserve Board’s Flow of Funds Accounts (FFA) provide statistics on a rich set of credit market instruments. However, it is challenging to inter pret such data using economic models that speak to the allocation of risk across agents, such as households or intermediaries. On the one hand, an instrument class such as “Treasury bonds” typically contains many dif ferent instruments that trade at different prices and have different exposure to interest rate shocks (for example, because of differences in duration). On the other hand, a lot of the price movements in instruments like Treasury bonds and mortgage backed securities are due to com mon interest rate shocks, making those instru ments close substitutes from a portfolio choice perspective. For understanding how interest rate risk is allocated in the economy, one would thus like to use information on many positions at the same time, rather than, say, focus on one set of instruments only. At the same time, models with many closely substitutable assets are problem atic. Instead, it would be desirable to compress position data into simple sets of portfolios, like “long” and “short” bonds, but with some con fidence that the risk properties of the original instruments are not lost along the way. DemanD anD Supply for Government BonDS