6.8.1 Liquidity Risk & Asset Pricing from Zero

Basic Prerequisites: none beyond basic algebra.

1. Intuition & Practical Objective

Start with the question this whole folder answers: why would two securities with identical future cash flows trade at different prices?

The answer is the cost of exiting. A government bond you can sell into the middle of the quote in a millisecond for $99.90 costs you almost nothing to turn back into cash. A private loan or a small-cap stock that takes a week to sell, and whose sale itself pushes the price down, is worth less to you today - because when you eventually want the cash back, it will not be there for free. Rational investors demand to be paid for that: they bid the illiquid asset down until its expected return is high enough to compensate for the round-trip cost.

That extra required return is the liquidity premium. Three ideas, in increasing order of subtlety:

The level of illiquidity is priced (Amihud–Mendelson). More illiquid assets earn higher expected returns, roughly the spread amortized over the holding period: the less often you trade, the more the one-time cost can be spread out.
The risk of illiquidity is priced too (Pástor–Stambaugh). Beyond the average level, investors fear changes in liquidity. A stock that tends to fall precisely when the whole market’s liquidity dries up is riskier and earns a premium - this is a liquidity beta, like a market beta but for liquidity instead of the market.
Illiquidity is common and it spikes in bad times (Acharya–Pedersen; Bao et al.). Illiquidity across assets moves together, and it worsens exactly in down markets - which is why the risk is so hard to hedge and so expensive.

One mental model to keep. Liquidity is insurance with a price. In calm markets you almost never collect; in a crisis it is the difference between getting out at 99 and getting out at 60. The premium you see in normal expected returns is the market charging you, in advance, for the states where you most need to trade and can least afford to.

This folder is the asset-pricing view. The sibling folder Liquidity Risk & Funding (Pillar 4) asks the funding question - can you finance the position at all? - and its L-VaR and margin-spiral machinery. Here we ask: given the cost of trading, what must prices and expected returns be? They are two views of one phenomenon; a full picture needs both.

2. Mathematical Ground Truth & Derivations

The amortization identity (Amihud–Mendelson 1986). Suppose the one-way cost of trading asset $i$ is a fraction $s_{i}$ of its value (the relative half-spread, the standard Foucault convention). A round trip (buy then sell) costs $2 s_{i}$ ; over a holding period of $h$ years that is roughly $2 s_{i} / h$ per year - which the literature usually writes as $s / h$ with $s$ instead denoting the relative spread (so $s = 2 \times$ the one-way cost). Here $s_{i}$ is the one-way cost. To make the asset competitive with a liquid asset returning $r$ , its gross required return must be

R_{i} ≃ r + \frac{s _{i}}{h},

where the precise constant depends on how the two sides of the cost are counted (Foucault et al. §9; a common form quotes the premium as $s / h$ with $s$ the one-way relative cost). The premium shrinks as the holding period grows - long-horizon investors can amortize a fixed cost over more time, so they demand less compensation. This is the classic rationale for why long-horizon (pension) money tolerates illiquid assets.

Why the level shows up in prices. If asset $i$ is illiquid (large $s_{i}$ ) and asset $j$ is liquid, the price of $i$ must be lower today so that, over the holding period, its expected return is the liquid rate plus the amortized cost. Equivalently, in a discount-factor language, illiquid assets are priced by a lower valuation because their future cash flows are worth less once you subtract what it costs to convert them into cash.

Why the risk shows up too. Now let liquidity vary over time. If aggregate liquidity $L_{t}$ can drop, then an asset whose payoff $r_{t}^{i}$ is negatively correlated with $L_{t}$ (it pays off badly exactly when trading gets expensive) is an asset that fails you in the states you need it. Risk-averse investors require compensation for holding it - the liquidity beta channel of 03 · Liquidity as a Priced Factor. This is a second, independent premium over and above the level effect.

3. Computational Implementation - the premium in numbers

The level effect, quantified with the Amihud–Mendelson amortization. The print-verified output shows the whole point: longer holding periods ⇒ smaller required premium for the same spread.

A 20 bp cost is nothing if you hold for a year (0.2%/yr) but a huge 10% per year if you hold for a week. This is why liquidity premia are widest in assets you trade often or must exit fast, and why illiquidity is mostly a short-horizon phenomenon - exactly the reasoning the empirical cross-section confirms (small caps, bonds, private assets).

4. Failure Modes & First-Principles Breakdowns

The premium is not a constant. It depends on the holding period, the expected future spread (not just today’s), and the state of the world. Quoting “the” liquidity premium of an asset without the horizon is meaningless.
Level vs. risk are conflated. A stock can be illiquid on average (big level premium) but have low liquidity risk (stable liquidity) - the two channels of §2 must be estimated separately or the premium is double-counted.
Prices already embed the premium. You cannot “harvest” the liquidity premium by buying illiquid assets and hoping; part of the premium is compensation for bearing the risk that liquidity evaporates when you need out - see 05 · Failure Modes and the funding-side spiral.

5. References

Amihud & Mendelson (1986). Asset pricing and the bid-ask spread. Journal of Financial Economics 17, 223–249. The origin of the liquidity-premium idea; the spread/holding-period relation.
Amihud (2002). Illiquidity and stock returns. JFM 5, 31–56. The empirical anchor: expected illiquidity is priced cross-sectionally and over time.
Amihud, Mendelson & Pedersen (2013). Market Liquidity: Asset Pricing, Risk, and Crises. Cambridge University Press
Foucault, Pagano & Roell (2013). Market Liquidity.

6. Connected Graph Bridges

Next: 02 · Illiquidity Measures · 03 · Liquidity as a Priced Factor
Back: Index Hub
Sibling funding view: Liquidity Risk & Funding (Pillar 4) · Market Impact & Depth

Kwant Atlas

Explorer