Corporate Bond Liquidity And Dealer Balance Sheets

Buy a share of a large-cap stock and you're trading against a continuous, anonymous order book with a market maker or another investor on the other side within milliseconds. Buy a corporate bond and there's usually no order book at all: you call or message a dealer, who quotes you a price out of their own inventory or agrees to source the bond from another dealer. That structural difference means corporate bond liquidity isn't really about order flow the way equity liquidity is — it's about how much of their own capital dealers are willing to put at risk holding bonds on their books.

The analogy: a wine merchant's cellar space

A wine merchant doesn't have infinite cellar space. When a customer wants to sell a case of an obscure vintage, the merchant has to decide whether to buy it outright, tying up cellar space and capital until another customer wants that exact vintage, or decline and point the seller elsewhere. When cellar space is plentiful and cheap, the merchant buys generously and quotes tight, welcoming prices. When space is tight — after a big purchase, or when the merchant is nervous about a glut — every additional case competes for scarce room, and quotes widen or the merchant simply stops buying. Bond dealers behave the same way with balance sheet: each bond they hold in inventory consumes capital, uses up regulatory capacity, and adds to risk they must fund, so their willingness to make markets rises and falls with how much "cellar space" they have free.

Why this shows up as liquidity that vanishes in stress

Post-2008 capital regulations (Basel III's leverage ratio, in particular) made holding bond inventory materially more expensive for bank-affiliated dealers, and this cost is roughly proportional to balance sheet size regardless of how safe an individual bond is. The result is that dealer bond inventories have shrunk structurally relative to the size of the outstanding bond market, so the market has less capacity to absorb one-sided selling than it once did.

$$\text{Bid-ask spread on a bond} \approx f(\text{credit risk}) + g(\text{dealer balance sheet cost})$$

In plain English: the price you pay to trade a bond reflects not just how risky the issuer is, but a separate premium for how expensive it currently is for a dealer to hold that bond on their books — and that second term can move sharply even when nothing about the issuer's credit has changed.

Worked example: a selloff widening spreads with no credit news

A BBB-rated corporate bond normally trades with a 10-basis-point bid-ask spread. A broad market selloff hits, unrelated to this specific issuer — perhaps a rates shock or a wave of mutual fund redemptions across the sector. Dealers across the Street are simultaneously being asked to buy bonds from clients who need cash, and their balance sheets are filling up with inventory they didn't want. A dealer who would normally quote 99.50/99.60 for this bond might now quote 98.80/99.40 — an 60-cent spread, six times wider — not because anyone thinks the issuer is more likely to default, but because the dealer's capital is scarcer and every additional bond taken onto the book now carries a higher opportunity cost. If several clients try to sell $5 million each simultaneously, the third and fourth callers may get materially worse prices than the first simply because the dealer's remaining balance sheet capacity has been used up by the first two trades.

What this means in practice

Corporate bond liquidity should be read as a joint signal of credit conditions and dealer capacity, not credit risk alone — spreads can blow out during a funding squeeze even for issuers with stable fundamentals, and they can compress even for shakier issuers when dealers have ample capital to deploy. Traders sizing bond positions need to think about how the trade competes for balance sheet against everything else the dealer is being asked to do that day, especially in stress periods when many sellers show up to the same handful of dealers at once.