Findings
Four findings about who absorbs Treasury pressure.
Treasury resilience depends on which investors adjust, which maturities reach the market, and whether demand changes are expected to persist.
Finding 01 · Supply absorption
Different investors absorbed different parts of the supply expansion.
Select a period and maturity to identify the largest increase and reduction in observed holdings.
Federal Reserve
−405.1 $bnObserved holdings
Interpretation: The post-2022 bill expansion was absorbed especially strongly by money market funds, while medium- and long-maturity changes involved different sectors.
Source: Dissecting Treasury Market Resilience, Figure 9. Holdings changes are descriptive accounting, not causal estimates of why a sector bought or sold.
Finding 02 · Maturity
Dealers and hedge funds sharply limit price impact in the T-bill market.
Own-maturity price impact is the percentage change in prices of Treasuries in a maturity bucket after a latent demand shock to that same bucket. Each shock equals 1% of the bucket’s outstanding supply; price changes in other maturity buckets are excluded. For the T-bill bucket (less than one year), removing dealers and hedge funds makes this price impact 75.2 times larger.
What this tab shows: A latent demand change equal to 1% of T-bills outstanding hits the T-bill bucket. The measure is the resulting percentage change in T-bill prices only.
With arbitrageurs (dealers and hedge funds)1.0×
Without arbitrageurs75.2×
75.2× larger own-maturity price impact without intermediaries, revealing their especially large role in the T-bill market.
Model comparison
Interpretation: Removing arbitrageurs increases the T-bill own-maturity price impact 75.2 times, compared with 9.3 times at intermediate maturities and 3.5 times at long maturities. Their price-stabilizing role is therefore largest in the T-bill market, where they can absorb imbalances at near-zero duration cost.
Source: Granular Treasury Demand with Arbitrageurs, Table 4, “Price Impact of Latent Demand Shocks with and without Arbitrageurs.” Each latent-demand shock equals 1% of outstanding supply in one maturity bucket. Panel (a) includes arbitrageurs, Panel (b) removes them, and Panel (c) divides Panel (b) by Panel (a). The displayed ratios are Panel (c)’s diagonal elements, which measure the price change in the same maturity bucket as the shock and exclude price responses at other maturities. Short maturity is less than one year. Sample: 2011Q4–2022Q4.
Finding 03 · Foreign demand
Equal foreign withdrawals do not have equal yield effects.
Select a reported location to compare its maturity mix and model-implied 10-year response.
Reported location
Canada
Estimated steady-state holdings$174.5bn
10-year effect of −$10bn+0.67 bp
Maturity mix22% / 49% / 29%short / medium / long
Model counterfactual
Interpretation: Holding the withdrawal at $10 billion isolates differences in estimated demand schedules and maturity composition rather than country size.
Source: Dissecting Treasury Market Resilience, Figure 3 and Appendix Tables A5–A6. Results are point estimates for 34 sufficiently large locations; TIC labels reflect reporting residence and custody.
Finding 04 · Future demand
At longer maturities, expected future support can matter more than today’s balance-sheet change.
Compare passive Fed runoff with weaker expected Fed support as yields and economic conditions change.
Passive runoff0.49 bp
Weaker expected Fed support0.09 bp
The immediate holdings reduction is the larger channel at this maturity.
Policy counterfactual
Interpretation: Immediate runoff matters more at five years. Farther out the curve, changing expected future Fed demand becomes increasingly important.
Source: Dissecting Treasury Market Resilience, Figure 10. The policy paths are model scenarios rather than estimates of historical beliefs.