economicsslides
TRANSCRIPT
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
The OTC Theory of the Fed Funds Market:A Primer
Gara Afonso Ricardo Lagos
FRB of New York New York University
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
The market for federal funds
A market for loans of reserve balances at the Fed.
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
The market for federal funds
What’s traded?
Unsecured loans (mostly overnight)
How are they traded?
Over the counter
Who trades?
Commercial banks, securities dealers, agencies and branches offoreign banks in the U.S., thrift institutions, federal agencies
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
The market for federal funds
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Why is the fed funds market interesting?
It is an interesting example of an OTC market
(Unusually good data is available)
Reallocates reserves among banks
(Banks use it to offset liquidity shocks and manage reserves)
Determines the interest rate on the shortest maturityinstrument in the term structure
Is the “epicenter”of monetary policy implementation
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Why is the fed funds market interesting?
It is an interesting example of an OTC market
(Unusually good data is available)
Reallocates reserves among banks
(Banks use it to offset liquidity shocks and manage reserves)
Determines the interest rate on the shortest maturityinstrument in the term structure
Is the “epicenter”of monetary policy implementation
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Why is the fed funds market interesting?
It is an interesting example of an OTC market
(Unusually good data is available)
Reallocates reserves among banks
(Banks use it to offset liquidity shocks and manage reserves)
Determines the interest rate on the shortest maturityinstrument in the term structure
Is the “epicenter”of monetary policy implementation
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Why is the fed funds market interesting?
It is an interesting example of an OTC market
(Unusually good data is available)
Reallocates reserves among banks
(Banks use it to offset liquidity shocks and manage reserves)
Determines the interest rate on the shortest maturityinstrument in the term structure
Is the “epicenter”of monetary policy implementation
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
In this paper we ...
(1) Propose an OTC model of trade in the fed funds market
(2) Use the theory to address some elementary questions, e.g.,
Positive: How is the fed funds rate determined?
Normative: Is the OTC market structure able to achieve aneffi cient reallocation of funds?
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
The model
A trading session in continuous time, t ∈ [0,T ], τ ≡ T − t
Unit measure of banks hold reserve balances k (τ) ∈ {0, 1, 2}
{nk (τ)} : distribution of balances at time T − τ
Linear payoffs from balances, discount at rate r
Uk : payoff from holding balance k at the end of the session
Trade opportunities are bilateral and random (Poisson rate α)
Loan and repayment amounts determined by Nash bargaining
Assume all loans repaid at time T + ∆, where ∆ ∈ R+
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Institutional features of the fed funds market
Model Fed funds market
Search and bargaining
Over-the-counter market
[0,T ] 4:00pm-6:30pm
{nk (T )}Distribution of reservebalances at 4:00pm
{Uk} Reserve requirements,interest on reserves...
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Institutional features of the fed funds market
Model Fed funds market
Search and bargaining Over-the-counter market
[0,T ] 4:00pm-6:30pm
{nk (T )}Distribution of reservebalances at 4:00pm
{Uk} Reserve requirements,interest on reserves...
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Institutional features of the fed funds market
Model Fed funds market
Search and bargaining Over-the-counter market
[0,T ]
4:00pm-6:30pm
{nk (T )}Distribution of reservebalances at 4:00pm
{Uk} Reserve requirements,interest on reserves...
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Institutional features of the fed funds market
Model Fed funds market
Search and bargaining Over-the-counter market
[0,T ] 4:00pm-6:30pm
{nk (T )}Distribution of reservebalances at 4:00pm
{Uk} Reserve requirements,interest on reserves...
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Institutional features of the fed funds market
Model Fed funds market
Search and bargaining Over-the-counter market
[0,T ] 4:00pm-6:30pm
{nk (T )}
Distribution of reservebalances at 4:00pm
{Uk} Reserve requirements,interest on reserves...
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Institutional features of the fed funds market
Model Fed funds market
Search and bargaining Over-the-counter market
[0,T ] 4:00pm-6:30pm
{nk (T )}Distribution of reservebalances at 4:00pm
{Uk} Reserve requirements,interest on reserves...
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Institutional features of the fed funds market
Model Fed funds market
Search and bargaining Over-the-counter market
[0,T ] 4:00pm-6:30pm
{nk (T )}Distribution of reservebalances at 4:00pm
{Uk}
Reserve requirements,interest on reserves...
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Institutional features of the fed funds market
Model Fed funds market
Search and bargaining Over-the-counter market
[0,T ] 4:00pm-6:30pm
{nk (T )}Distribution of reservebalances at 4:00pm
{Uk} Reserve requirements,interest on reserves...
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Bellman equations
rV1 (τ) + V1 (τ) = 0
rV0 (τ) + V0 (τ) = αn2 (τ)max {V1 (τ)− V0 (τ)− R (τ) , 0}
rV2 (τ) + V2 (τ) = αn0 (τ)max {V1 (τ)− V2 (τ) + R (τ) , 0}
Vk (0) = Uk
R (τ) = argmaxR[V1 (τ)− R − V0 (τ)]θ [V1 (τ) + R − V2 (τ)]1−θ
= θ [V2 (τ)− V1 (τ)] + (1− θ) [V1 (τ)− V0 (τ)]
R (τ) ≡ e−r (τ+∆)R (τ) (PDV of repayment)
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Bellman equations
rV1 (τ) + V1 (τ) = 0
rV0 (τ) + V0 (τ) = αn2 (τ)max {V1 (τ)− V0 (τ)− R (τ) , 0}
rV2 (τ) + V2 (τ) = αn0 (τ)max {V1 (τ)− V2 (τ) + R (τ) , 0}
Vk (0) = Uk
R (τ) = argmaxR[V1 (τ)− R − V0 (τ)]θ [V1 (τ) + R − V2 (τ)]1−θ
= θ [V2 (τ)− V1 (τ)] + (1− θ) [V1 (τ)− V0 (τ)]
R (τ) ≡ e−r (τ+∆)R (τ) (PDV of repayment)
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Bellman equations
rV1 (τ) + V1 (τ) = 0
rV0 (τ) + V0 (τ) = αn2 (τ) φ (τ) θS (τ)
rV2 (τ) + V2 (τ) = αn0 (τ) φ (τ) (1− θ) S (τ)
where:
S (τ) ≡ 2V1 (τ)− V0 (τ)− V2 (τ)
φ (τ) =
{1 if 0 < S (τ)0 if S (τ) ≤ 0
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Time-path for the distribution of balances
n0 (τ) = αφ (τ) n2 (τ) n0 (τ)
n1 (τ) = −2αφ (τ) n2 (τ) n0 (τ)
n2 (τ) = αφ (τ) n2 (τ) n0 (τ)
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
DefinitionAn equilibrium is a value function, V, a path for the distribution ofreserve balances, n (τ), and a path for the distribution of tradingprobabilities, φ (τ), such that:
(a) given the value function and the distribution of tradingprobabilities, the distribution of balances evolves according to thelaw of motion; and
(b) given the path for the distribution of balances, the valuefunction and the distribution of trading probabilities satisfyindividual optimization given the bargaining protocol.
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Analysis
S (τ) = −δ (τ) S (τ)
δ (τ) ≡ {r + αφ (τ) [θn2 (τ) + (1− θ) n0 (τ)]}
⇒
S (τ) = e−δ(τ)S (0)
δ (τ) ≡∫ τ
0δ (x) dx
S (τ) ≡ 2V1 (τ)− V0 (τ)− V2 (τ)
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Analysis
S (τ) = −δ (τ) S (τ)
δ (τ) ≡ {r + αφ (τ) [θn2 (τ) + (1− θ) n0 (τ)]}
⇒
S (τ) = e−δ(τ)S (0)
δ (τ) ≡∫ τ
0δ (x) dx
S (τ) ≡ 2V1 (τ)− V0 (τ)− V2 (τ)
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Assumption: S (0) ≡ 2U1 − U2 − U0 > 0
⇒
S (τ) = e−δ(τ)S (0) > 0 for all τ
⇒
φ (τ) = 1 for all τ
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Assumption: S (0) ≡ 2U1 − U2 − U0 > 0
⇒
S (τ) = e−δ(τ)S (0) > 0 for all τ
⇒
φ (τ) = 1 for all τ
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Assumption: S (0) ≡ 2U1 − U2 − U0 > 0
⇒
S (τ) = e−δ(τ)S (0) > 0 for all τ
⇒
φ (τ) = 1 for all τ
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Proposition
Suppose 2U1 − U2 − U0 > 0. Then the unique equilibrium is:
n2 (τ) = n0 (τ) + n2 (T )− n0 (T )n1 (τ) = 1− 2n0 (τ) + n0 (T )− n2 (T )n0 (τ) = [n2(T )−n0(T )]n0(T )
eα[n2 (T )−n0 (T )](T−τ)n2(T )−n0(T )
S (τ) =[n2(T )−e−α[n2 (T )−n0 (T )](T−τ)n0(T )]e−{r+αθ[n2 (T )−n0 (T )]}τ
n2(T )−e−α[n2 (T )−n0 (T )]T n0(T )S (0)
R (τ) = er∆ {β (τ) (U2 − U1) + [1− β (τ)] (U1 − U0)}
β (τ) ≡ θ[n2(T )−e−α[n2 (T )−n0 (T )](T−τ)n0(T )]e−αθ[n2 (T )−n0 (T )]τ
n2(T )−e−α[n2 (T )−n0 (T )]T n0(T )
+[1−e−αθ[n2 (T )−n0 (T )]τ]n2(T )n2(T )−e−α[n2 (T )−n0 (T )]T n0(T )
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Proposition
Suppose 2U1 − U2 − U0 > 0. Then the unique equilibrium is:
n2 (τ) = n0 (τ) + n2 (T )− n0 (T )n1 (τ) = 1− 2n0 (τ) + n0 (T )− n2 (T )n0 (τ) = [n2(T )−n0(T )]n0(T )
eα[n2 (T )−n0 (T )](T−τ)n2(T )−n0(T )
S (τ) =[n2(T )−e−α[n2 (T )−n0 (T )](T−τ)n0(T )]e−{r+αθ[n2 (T )−n0 (T )]}τ
n2(T )−e−α[n2 (T )−n0 (T )]T n0(T )S (0)
R (τ) = er∆ {β (τ) (U2 − U1) + [1− β (τ)] (U1 − U0)}
β (τ) ≡ θ[n2(T )−e−α[n2 (T )−n0 (T )](T−τ)n0(T )]e−αθ[n2 (T )−n0 (T )]τ
n2(T )−e−α[n2 (T )−n0 (T )]T n0(T )
+[1−e−αθ[n2 (T )−n0 (T )]τ]n2(T )n2(T )−e−α[n2 (T )−n0 (T )]T n0(T )
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Proposition
Suppose 2U1 − U2 − U0 > 0. Then the unique equilibrium is:
n2 (τ) = n0 (τ) + n2 (T )− n0 (T )n1 (τ) = 1− 2n0 (τ) + n0 (T )− n2 (T )n0 (τ) = [n2(T )−n0(T )]n0(T )
eα[n2 (T )−n0 (T )](T−τ)n2(T )−n0(T )
S (τ) =[n2(T )−e−α[n2 (T )−n0 (T )](T−τ)n0(T )]e−{r+αθ[n2 (T )−n0 (T )]}τ
n2(T )−e−α[n2 (T )−n0 (T )]T n0(T )S (0)
R (τ) = er∆ {β (τ) (U2 − U1) + [1− β (τ)] (U1 − U0)}
β (τ) ≡ θ[n2(T )−e−α[n2 (T )−n0 (T )](T−τ)n0(T )]e−αθ[n2 (T )−n0 (T )]τ
n2(T )−e−α[n2 (T )−n0 (T )]T n0(T )
+[1−e−αθ[n2 (T )−n0 (T )]τ]n2(T )n2(T )−e−α[n2 (T )−n0 (T )]T n0(T )
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Proposition
Suppose 2U1 − U2 − U0 > 0. Then the unique equilibrium is:
n2 (τ) = n0 (τ) + n2 (T )− n0 (T )n1 (τ) = 1− 2n0 (τ) + n0 (T )− n2 (T )n0 (τ) = [n2(T )−n0(T )]n0(T )
eα[n2 (T )−n0 (T )](T−τ)n2(T )−n0(T )
S (τ) =[n2(T )−e−α[n2 (T )−n0 (T )](T−τ)n0(T )]e−{r+αθ[n2 (T )−n0 (T )]}τ
n2(T )−e−α[n2 (T )−n0 (T )]T n0(T )S (0)
R (τ) = er∆ {β (τ) (U2 − U1) + [1− β (τ)] (U1 − U0)}
β (τ) ≡ θ[n2(T )−e−α[n2 (T )−n0 (T )](T−τ)n0(T )]e−αθ[n2 (T )−n0 (T )]τ
n2(T )−e−α[n2 (T )−n0 (T )]T n0(T )
+[1−e−αθ[n2 (T )−n0 (T )]τ]n2(T )n2(T )−e−α[n2 (T )−n0 (T )]T n0(T )
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Proposition
Suppose 2U1 − U2 − U0 > 0. Then, the equilibrium supports aneffi cient allocation of reserve balances.
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intuition for effi ciency result
rV0 (τ) + V0 (τ) = αn2 (τ) θS (τ)
rλ0 (τ) + λ0 (τ) = αn2 (τ) S∗ (τ)
rV1 (τ) + V1 (τ) = 0
rλ1 (τ) + λ1 (τ) = 0
rV2 (τ) + V2 (τ) = αn0 (τ) (1− θ) S (τ)
rλ2 (τ) + λ2 (τ) = αn0 (τ) S∗ (τ)
S (τ) = e−δ(τ)S (0)
S∗ (τ) = e−δ∗(τ)S (0)
δ∗(τ)− δ (τ) = α
∫ τ
0[(1− θ) n2 (z) + θn0 (z)] dz ≥ 0
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intuition for effi ciency result
rV0 (τ) + V0 (τ) = αn2 (τ) θS (τ)
rλ0 (τ) + λ0 (τ) = αn2 (τ) S∗ (τ)
rV1 (τ) + V1 (τ) = 0
rλ1 (τ) + λ1 (τ) = 0
rV2 (τ) + V2 (τ) = αn0 (τ) (1− θ) S (τ)
rλ2 (τ) + λ2 (τ) = αn0 (τ) S∗ (τ)
S (τ) = e−δ(τ)S (0)
S∗ (τ) = e−δ∗(τ)S (0)
δ∗(τ)− δ (τ) = α
∫ τ
0[(1− θ) n2 (z) + θn0 (z)] dz ≥ 0
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intuition for effi ciency result
rV0 (τ) + V0 (τ) = αn2 (τ) θS (τ)
rλ0 (τ) + λ0 (τ) = αn2 (τ) S∗ (τ)
rV1 (τ) + V1 (τ) = 0
rλ1 (τ) + λ1 (τ) = 0
rV2 (τ) + V2 (τ) = αn0 (τ) (1− θ) S (τ)
rλ2 (τ) + λ2 (τ) = αn0 (τ) S∗ (τ)
S (τ) = e−δ(τ)S (0)
S∗ (τ) = e−δ∗(τ)S (0)
δ∗(τ)− δ (τ) = α
∫ τ
0[(1− θ) n2 (z) + θn0 (z)] dz ≥ 0
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intuition for effi ciency result
rV0 (τ) + V0 (τ) = αn2 (τ) θS (τ)
rλ0 (τ) + λ0 (τ) = αn2 (τ) S∗ (τ)
rV1 (τ) + V1 (τ) = 0
rλ1 (τ) + λ1 (τ) = 0
rV2 (τ) + V2 (τ) = αn0 (τ) (1− θ) S (τ)
rλ2 (τ) + λ2 (τ) = αn0 (τ) S∗ (τ)
S (τ) = e−δ(τ)S (0)
S∗ (τ) = e−δ∗(τ)S (0)
δ∗(τ)− δ (τ) = α
∫ τ
0[(1− θ) n2 (z) + θn0 (z)] dz ≥ 0
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intuition for effi ciency result
Equilibrium:
Gain from trade as perceived by borrower: θS (τ)
Gain from trade as perceived by lender: (1− θ) S (τ)
Planner:
Each of their marginal contributions equals S∗ (τ)
δ∗ (τ) ≥ δ (τ) for all τ ∈ [0,T ], with “=”only for τ = 0
⇒ The planner “discounts”more heavily than the equilibrium
⇒ S∗ (τ) < S (τ) for all τ ∈ (0, 1]⇒ Social value of loan < joint private value of loan
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intuition for effi ciency result
Equilibrium:
Gain from trade as perceived by borrower: θS (τ)
Gain from trade as perceived by lender: (1− θ) S (τ)
Planner:
Each of their marginal contributions equals S∗ (τ)
δ∗ (τ) ≥ δ (τ) for all τ ∈ [0,T ], with “=”only for τ = 0
⇒ The planner “discounts”more heavily than the equilibrium
⇒ S∗ (τ) < S (τ) for all τ ∈ (0, 1]⇒ Social value of loan < joint private value of loan
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intuition for effi ciency result
Equilibrium:
Gain from trade as perceived by borrower: θS (τ)
Gain from trade as perceived by lender: (1− θ) S (τ)
Planner:
Each of their marginal contributions equals S∗ (τ)
δ∗ (τ) ≥ δ (τ) for all τ ∈ [0,T ], with “=”only for τ = 0
⇒ The planner “discounts”more heavily than the equilibrium
⇒ S∗ (τ) < S (τ) for all τ ∈ (0, 1]⇒ Social value of loan < joint private value of loan
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intuition for effi ciency result
Equilibrium:
Gain from trade as perceived by borrower: θS (τ)
Gain from trade as perceived by lender: (1− θ) S (τ)
Planner:
Each of their marginal contributions equals S∗ (τ)
δ∗ (τ) ≥ δ (τ) for all τ ∈ [0,T ], with “=”only for τ = 0
⇒ The planner “discounts”more heavily than the equilibrium
⇒ S∗ (τ) < S (τ) for all τ ∈ (0, 1]⇒ Social value of loan < joint private value of loan
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intuition for effi ciency result
Equilibrium:
Gain from trade as perceived by borrower: θS (τ)
Gain from trade as perceived by lender: (1− θ) S (τ)
Planner:
Each of their marginal contributions equals S∗ (τ)
δ∗ (τ) ≥ δ (τ) for all τ ∈ [0,T ], with “=”only for τ = 0
⇒ The planner “discounts”more heavily than the equilibrium
⇒ S∗ (τ) < S (τ) for all τ ∈ (0, 1]⇒ Social value of loan < joint private value of loan
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intuition for effi ciency result
Equilibrium:
Gain from trade as perceived by borrower: θS (τ)
Gain from trade as perceived by lender: (1− θ) S (τ)
Planner:
Each of their marginal contributions equals S∗ (τ)
δ∗ (τ) ≥ δ (τ) for all τ ∈ [0,T ], with “=”only for τ = 0
⇒ The planner “discounts”more heavily than the equilibrium
⇒ S∗ (τ) < S (τ) for all τ ∈ (0, 1]⇒ Social value of loan < joint private value of loan
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intuition for effi ciency result
Planner internalizes that searching borrowers and lendersmake it easier for other lenders and borrowers to find partners
These “liquidity provision services” to others receive nocompensation in the equilibrium, so individual agents ignorethem when calculating their equilibrium payoffs
The equilibrium payoff to lenders may be too high or too lowrelative to their shadow price in the planner’s problem:
E.g., too high if (1− θ) S (τ) > S∗ (τ)
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intuition for effi ciency result
Planner internalizes that searching borrowers and lendersmake it easier for other lenders and borrowers to find partners
These “liquidity provision services” to others receive nocompensation in the equilibrium, so individual agents ignorethem when calculating their equilibrium payoffs
The equilibrium payoff to lenders may be too high or too lowrelative to their shadow price in the planner’s problem:
E.g., too high if (1− θ) S (τ) > S∗ (τ)
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intuition for effi ciency result
Planner internalizes that searching borrowers and lendersmake it easier for other lenders and borrowers to find partners
These “liquidity provision services” to others receive nocompensation in the equilibrium, so individual agents ignorethem when calculating their equilibrium payoffs
The equilibrium payoff to lenders may be too high or too lowrelative to their shadow price in the planner’s problem:
E.g., too high if (1− θ) S (τ) > S∗ (τ)
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Frictionless limit
Proposition
Let Q ≡ ∑2k=0 knk (T ) = 1+ n2 (T )− n0 (T ).
For τ ∈ [0,T ], as α→ ∞, ρ (τ)→ ρ∞, where
1+ ρ∞ =
er∆ (U1 − U0) if Q < 1er∆ [θ (U2 − U1) + (1− θ) (U1 − U0)] if Q = 1er∆ (U2 − U1) if 1 < Q.
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Positive implications
The theory delivers:
(1) Time-varying trade volume
(2) Time-varying fed fund rate
(3) Intraday convergence of distribution of reserve balances
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Trade volume
Flow volume of trade at time T − τ:
υ (τ) = αn0 (τ) n2 (τ)
Total volume traded during the trading session:
υ =∫ T
0υ (τ) dτ
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Fed funds rateThe gross rate on a loan at time T − τ is:
1+ ρ (τ) = er∆ {β (τ) (U2 − U1) + [1− β (τ)] (U1 − U0)}
withβ (τ) ∈ [0, 1]
The average daily rate is:
ρ =1T
∫ T
0ρ (τ) dτ
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Cross-sectional distribution of reserve balances
Let µ (τ) and σ2 (τ) denote the mean and variance of thecross-sectional distribution of reserve balances at t = T − τ
µ (τ) = 1+ n2 (T )− n0 (T ) ≡ Q
σ2 (τ) = σ2 (T )− 2 [2+ n2 (T )− n0 (T )] [n0 (T )− n0 (τ)]
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
U1 = e−r∆(1+ i rf )
U2 = e−r∆(2+ i rf + ief )
U0 = −e−r∆(iwf − i rf + Pw )
Proposition
ρ (τ) = β (τ) ief + [1− β (τ)] (iwf + Pw ) where
1 If n2 (T ) = n0 (T ), β (τ) = θ
2 If n2 (T ) < n0 (T ), β (τ) ∈ [0, θ], β (0) = θ and β′ (τ) < 0
3 If n0 (T ) < n2 (T ), β (τ) ∈ [θ, 1], β (0) = θ and β′ (τ) > 0.
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
U1 = e−r∆(1+ i rf )
U2 = e−r∆(2+ i rf + ief )
U0 = −e−r∆(iwf − i rf + Pw )
Proposition
ρ (τ) = β (τ) ief + [1− β (τ)] (iwf + Pw ) where
1 If n2 (T ) = n0 (T ), β (τ) = θ
2 If n2 (T ) < n0 (T ), β (τ) ∈ [0, θ], β (0) = θ and β′ (τ) < 0
3 If n0 (T ) < n2 (T ), β (τ) ∈ [θ, 1], β (0) = θ and β′ (τ) > 0.
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intraday interest rate in a balanced market
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intraday interest rate in a market with shortage of reserves
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intraday interest rate in a market with excess reserves
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Interest rate and the quantity of reserves
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
k = 1
Two scenarios{nH0 (T ) , n
L2 (T )
} {nL0 (T ) , n
H2 (T )
}{0.6, 0.3} {0.3, 0.6}
Experiments
Bargaining Power (θ) Discount Rate (iwf ) Contact Rate (α)
0.1 0.5 0.9 .0050360
.0075360
.0100360 25 50 100
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Bargaining power
16:00 16:30 17:00 17:30 18:00 18:300
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6x 10 5
Sur
plus
Eastern Time
θ=0.1θ=0.5θ=0.9
16:00 16:30 17:00 17:30 18:00 18:301.000006
1.000008
1.00001
1.000012
1.000014
1.000016
1.000018
1.00002
V2V
1
Eastern Time
θ=0.1θ=0.5θ=0.9
16:00 16:30 17:00 17:30 18:00 18:30
0.3
0.4
0.5
0.6
0.7
0.8
ρ (%
)
Eastern Time
θ=0.1θ=0.5θ=0.9
16:00 16:30 17:00 17:30 18:00 18:300
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6x 10 5
Sur
plus
Eastern Time
θ=0.1θ=0.5θ=0.9
16:00 16:30 17:00 17:30 18:00 18:30
1.000007
1.0000075
1.000008
1.0000085
1.000009
V2V
1
Eastern Time
θ=0.1θ=0.5θ=0.9
16:00 16:30 17:00 17:30 18:00 18:30
0.3
0.4
0.5
0.6
0.7
0.8
ρ (%
)Eastern Time
θ=0.1θ=0.5θ=0.9
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Contact rate
16:00 16:30 17:00 17:30 18:00 18:300
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6x 10 5
Sur
plus
Eastern Time
α=25α=50α=100
16:00 16:30 17:00 17:30 18:00 18:301.000006
1.000008
1.00001
1.000012
1.000014
1.000016
1.000018
1.00002
V2V
1
Eastern Time
α=25α=50α=100
16:00 16:30 17:00 17:30 18:00 18:30
0.54
0.56
0.58
0.6
0.62
0.64
0.66
0.68
0.7
ρ (%
)
Eastern Time
α=25α=50α=100
16:00 16:30 17:00 17:30 18:00 18:300
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6x 10 5
Sur
plus
Eastern Time
α=25α=50α=100
16:00 16:30 17:00 17:30 18:00 18:30
1.000007
1.0000075
1.000008
1.0000085
1.000009
V2V
1
Eastern Time
α=25α=50α=100
16:00 16:30 17:00 17:30 18:00 18:30
0.35
0.4
0.45
0.5
0.55
ρ (%
)Eastern Time
α=25α=50α=100
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Discount-Window lending rate
16:00 16:30 17:00 17:30 18:00 18:300
0.5
1
1.5
2
2.5x 10 5
Sur
plus
Eastern Time
ifw=0.5%
ifw=0.75%
ifw=1%
16:00 16:30 17:00 17:30 18:00 18:30
1.000006
1.000008
1.00001
1.000012
1.000014
1.000016
1.000018
1.00002
V2V
1
Eastern Time
ifw=0.5%
ifw=0.75%
ifw=1%
16:00 16:30 17:00 17:30 18:00 18:30
0.4
0.45
0.5
0.55
0.6
0.65
0.7
0.75
0.8
ρ (%
)
Eastern Time
ifw=0.5%
ifw=0.75%
ifw=1%
16:00 16:30 17:00 17:30 18:00 18:300
0.5
1
1.5
2
2.5x 10 5
Sur
plus
Eastern Time
ifw=0.5%
ifw=0.75%
ifw=1%
16:00 16:30 17:00 17:30 18:00 18:30
1.000007
1.0000075
1.000008
1.0000085
1.000009
1.0000095
1.00001
1.0000105
V2V
1
Eastern Time
ifw=0.5%
ifw=0.75%
ifw=1%
16:00 16:30 17:00 17:30 18:00 18:30
0.3
0.35
0.4
0.45
0.5
0.55
0.6
0.65
0.7
ρ (%
)Eastern Time
ifw=0.5%
ifw=0.75%
ifw=1%
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
More to be done...
Fed funds brokers
Banks’portfolio decisions
Random “payment shocks”
Sequence of trading sessions
Ex-ante heterogeneity (αi , θi , U ik )
Generalized inventories
Quantitative work
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
The views expressed here are not necessarily reflective ofviews at the Federal Reserve Bank of New Yorkor the Federal Reserve System.
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Evidence of OTC frictions in the fed funds market
Price dispersion
Intermediation
Intraday evolution of the distribution of reserve balances
There are banks that are “very long”and buyThere are banks that are “very short” and sell
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Price dispersion
.15
.1
.05
0
.05
.1
Per
cent
16:00 16:30 17:00 17:30 18:00 18:30
10th/90th Percentiles 25th/75th Percentiles
Median Mean
Intraday Distribution of Fed Funds Spreads, 2005
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intermediation: excess funds reallocation
0
50
100
150
Bill
ions
of D
olla
rs
2005 2006 2007 2008 2009 2010
Excess Funds Reallocation
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intermediation: proportion of intermediated funds
0
.1
.2
.3
.4
.5
.6
2005 2006 2007 2008 2009 2010
Proportion of Intermediated Funds
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Intraday evolution of the distribution of reserve balances
.5
0
.5
1
16:00 16:30 17:00 17:30 18:00 18:30
10th/90th Percentiles 25th/75th Percentiles
Median Mean
Normalized Balances, 2007
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Banks that are “long”...and buy...
0
1,000
2,000
3,000
Milli
ons
of D
olla
rs
16:00 16:30 17:00 17:30 18:00 18:30
10th/90th Percentiles 25th/75th Percentiles
Median Mean
Purchases by Banks with Nonnegative Adjusted Balances, 2007
Intro Model Equilibrium Effi ciency Frictionless Implications Policy Examples Conclusion Appx.
Banks that are “short”...and sell...
0
100
200
300
400
500
Milli
ons
of D
olla
rs
16:00 16:30 17:00 17:30 18:00 18:30
10th/90th Percentiles 25th/75th Percentiles
Median Mean
Sales by Banks with Negative Adjusted Balances, 2007