Many economic questions require estimating the price impact of demand shifts (multipliers) in the bond market. Corporate bonds have salient characteristics that distinguish close versus distant substitutes. We show that accounting for the heterogeneous substitutability between bonds is critical for estimating multipliers correctly. By allowing for heterogeneous substitution, we find that security-level multipliers are essentially zero — an order of magnitude smaller than the estimate ignoring heterogeneous substitutability. Nonetheless, portfolio multipliers are substantially larger and monotonically increase with the aggregation level. Furthermore, we find that the multiplier is larger for high-yield bonds, longer-maturity bonds, and bonds with greater arbitrage risks.

This paper uses transaction-level data on UK inflation swaps to characterize who buys and sells inflation risk, when, and with what price elasticity. This provides measures of expected inflation cleaned from liquidity frictions, and of the varying influences of market participants with different beliefs. We first show that this market is segmented: pension funds trade at long horizons while hedge funds trade at short horizons, with dealer banks as their counterparties in both markets. This segmentation suggests three identification strategies — sign restrictions, granular instrumental variables, and heteroskedasticity — for the demand and supply functions of each investor type. We find that swap prices absorb new information quickly, the supply of long-horizon inflation protection is very elastic, short-horizon price movements are unreliable measures of expected inflation as they primarily reflect liquidity shocks, and that long-horizon price movements overstates changes in expected inflation during important events if they were not cleaned from liquidity shocks.

A growing literature uses portfolio holdings data to quantify the impact of investor demand on equilibrium prices via counterfactual experiments. The key parameter in relating demand and equilibrium prices is investors’ elasticity of demand with respect to the price. Unlike previous studies, which rely on cross-sectional estimates in levels, this paper proposes estimating elasticities from investors’ trades, that is changes in their portfolios. I use demand shocks from mutual fund flows as an instrument to address the endogeneity of trades and prices. Using the estimation in changes along with the flow-based instrument I find that elasticities are 4 times larger than what previous estimates suggest. Estimation over different trading horizons furthermore shows that investors become more elastic in the long run. The results suggest that the impact of demand shocks on equilibrium prices is smaller than previously estimated and partly reverts over time.