Streaming Prices Versus Request-For-Quote

Trade a liquid stock and you interact with a continuous, public order book — anyone can see the best bid and offer at any instant, and anyone can trade against them immediately. Trade an off-the-run corporate bond, or a large or unusual derivative, and that model mostly doesn't exist. Instead, you send a specific request to one or more dealers — "what's your price to buy $5 million of this bond?" — and only then, once they respond, do you have a tradeable price at all. These are the two fundamentally different ways a market can be organized: streaming and request-for-quote (RFQ).

The two models, side by side

	Streaming	RFQ
Price visibility	Continuous, public, visible before you trade	Only revealed after you ask, and usually only to you
Who sees your interest	Nobody, until your order is in the book or executes	The dealer(s) you asked, before you've committed to trade
Best suited to	Liquid, standardized instruments traded constantly	Illiquid, large, or customized instruments traded rarely
Information leakage	Low per-order (orders are typically small relative to the book)	Higher — asking for a quote signals your interest to whoever you ask

Why illiquid markets default to RFQ

A stock exchange can stream continuous prices because there's constant two-sided interest — someone is always willing to be on the other side at some price, and posting a firm, continuous quote doesn't cost a market maker much because the next trade is never far away. A dealer in a thinly traded corporate bond doesn't have that luxury: if they streamed a firm, tradeable two-way price on every bond they cover, all day, they'd be picked off constantly by counterparties who know something they don't, with no offsetting flow to average the risk away. RFQ lets the dealer price each individual trade on its own terms — checking their current inventory, current risk appetite, and the specific size being asked about — rather than committing to a price in advance for a trade that might not come for hours.

Walking through an RFQ

An asset manager wants to sell $10 million of a corporate bond. They send a request to three dealers simultaneously: "bid, $10mm, this CUSIP." Each dealer checks their own inventory and risk limits and responds within a minute or two:

Dealer	Bid
Dealer A	98.50
Dealer B	98.62
Dealer C	98.55

The asset manager sees all three bids and trades with Dealer B at 98.62 — the best price offered. Crucially, none of this was visible to anyone outside the request before it happened; the other two dealers now know they lost the trade, but the wider market saw nothing until (if at all) the trade is later reported.

The mechanics of running an RFQ well — how many dealers to ask, and whether to reveal your identity — are covered in Running An RFQ: How Many Dealers To Ask and All-To-All And Anonymous RFQ; RFQ And Request-For-Quote Execution covers the messaging standards that carry these requests and responses electronically.

The information-leakage trade-off inside RFQ itself

Even within RFQ, there's a further choice that matters a lot in practice: how many dealers do you ask? Asking only one dealer minimizes information leakage — nobody else learns you're trying to trade — but gives up any competitive pressure on price, since that one dealer knows they're the only one being asked and can price accordingly. Asking ten dealers creates real price competition, but now ten firms know a specific market participant is trying to sell a specific size of a specific instrument, and the nine who don't win the trade still walk away with that information, which they may use in ways that move the market against the eventual trade, or against the next one. Finding the right number to ask — enough for genuine competition, few enough to limit what leaks — is itself a live decision every RFQ desk makes trade by trade, and it's the central trade-off Running An RFQ: How Many Dealers To Ask works through in more detail.

Electronic RFQ platforms have also pushed some markets toward a hybrid: a request goes to a wider, sometimes anonymous pool of potential counterparties rather than a hand-picked few, trying to capture streaming-like competitive pricing while retaining RFQ's core property that a firm, tradeable price only exists once someone has actually asked for it.