Running An RFQ: How Many Dealers To Ask
Asking more dealers for a quote should mean more competition and a better price — except every dealer you ask also learns you're trading, which is why choosing how many (and which) dealers to include is a real trade-off, not just "ask everyone."
Prerequisites: Streaming Prices Versus Request-For-Quote
In markets that trade by request-for-quote — many corporate bonds, FX forwards, interest rate swaps — a client doesn't just see a price and hit a button; they send a request to a chosen set of dealers, each dealer responds with a price, and the client picks the best one. The obvious instinct is to ask as many dealers as possible: more competing quotes should mean a tighter price. In practice, the number and choice of dealers is a genuine trade-off, not a "more is always better" decision.
Why asking more dealers isn't free
Every dealer included in an RFQ learns something: that this specific client wants to trade this specific instrument, in roughly this size, right now. A dealer who doesn't win the trade still keeps that information — and if the trade is large or the instrument thin, several dealers now know a big order is coming, which can move the market against the client even before they've traded, or let a dealer who lost the RFQ trade ahead of the client's follow-up activity. This is information leakage, and it scales with the number of dealers asked, not just the number who win.
There's also a relationship cost: dealers who are asked to quote repeatedly and never win the trade eventually notice they're being used purely as a check on their competitors' pricing ("last look" fodder), and may respond by quoting worse prices to that client over time, or deprioritizing their requests, since a dealer's own resources spent quoting a trade they never win are a loss for them too.
The trade-off in practice
For a small, liquid trade, asking three to five dealers usually captures most of the competitive benefit with limited leakage, since the trade is easy for any dealer to absorb without much market impact. For a large or illiquid trade, clients often narrow the RFQ to one or two trusted dealers, or use a name-blind protocol, precisely because the leakage risk from broadcasting a big order to many counterparties can cost more than the extra competition saves. Some clients also rotate which dealers they include over time, both to preserve relationships and to avoid any one dealer building too complete a picture of the client's typical trading patterns.
Widening an RFQ to more dealers increases price competition but also increases information leakage about a client's trading intentions — the right number of dealers to ask depends on the trade's size and the instrument's liquidity, not a fixed rule of "ask everyone."
Treating "always ask the maximum number of dealers" as a free way to get the best price ignores the leakage cost entirely — for large or illiquid trades, the dealers who don't win can still move against you or front-run the information, which is exactly the cost a narrower, more selective RFQ is designed to avoid.
Related concepts
Practice in interviews
Further reading
- Harris, Trading and Exchanges, ch. 17
- Hendershott & Madhavan, Click or Call (2015)