Blog · Territory, capacity and quota planning · Trading
How a trading desk tiers its clients from the RFQ log and the blotter rather than from the coverage list: two axes, notional traded and hit ratio, four tiers that follow, the clients with high inquiry volume and low hit ratio who consume pricing effort for little, the clients with high hit ratio and modest flow who would trade more if shown more, and the quarterly re-tiering that moves clients on their own numbers.
A desk's client list is tiered by who the salespeople think matters. The RFQ log and the blotter can tier it by what each client actually does with the desk's prices: how much it trades and how much of what it asks for it takes. This guide sets out the two axes, the four tiers, what each tier means for coverage, and the re-tiering.
Per client, per product, per quarter:
Notional traded = Σ trades' notional Hit ratio, size-weighted = notional traded ÷ notional inquired
| Hit ratio high | Hit ratio low | |
|---|---|---|
| Flow high | Tier 1: protect. Axe first, salesperson's time | Tier 2: pricing review. The flow is there; the desk loses it |
| Flow low | Tier 3: show more. Trades what it sees | Tier 4: auto-price. Effort exceeds return |
Boundaries per product: a stated percentile of notional for high flow; a stated hit ratio, near the desk's median for the product, for high hit ratio.
Client identifiers only.
notional traded per client per product ≤ notional inquired Σ clients' notional traded = desk notional traded, per product
A client with trades and no inquiries in the product is listed; it is usually voice business not logged as RFQ, and it needs its inquiries recorded before the hit ratio means anything.
One product. High flow above $400m a quarter; high hit ratio above 35 percent.
| Client | Notional traded | Notional inquired | Hit ratio, size-weighted | Count-weighted | Tier | Instruction |
|---|---|---|---|---|---|---|
| 2207 | $1.4bn | $2.9bn | 48% | 52% | 1 | Protect |
| 4471 | $620m | $4.1bn | 15% | 41% | 2 | Pricing review: takes small, leaves large |
| 9034 | $180m | $310m | 58% | 55% | 3 | Show more |
| 1187 | $40m | $1.2bn | 3% | 9% | 4 | Auto-price |
Client 4471 trades six hundred million and asks for four billion, taking the small tickets and leaving the large. Tier two, with a pricing review on the large sizes. Client 1187 asks for a billion and trades forty million; a trader's time on its prices is not returning.
Quarterly, same two figures, same boundaries. Movements listed with the reason. A salesperson's flag carries a reason and a date and does not change the tier.
Tiers from the coverage list. Who is known, not who trades.
Count-weighted hit ratio. The client that leaves the large tickets looks fine.
Flow without hit ratio. The price-check client in tier one.
Voice business unlogged. The hit ratio for those clients is fiction.
Mapped once, the RFQ log, the blotter and the coverage file produce the two axes, the tiers, the instructions and the movements every quarter. Covirage builds this from the exports as they are. The trading page describes the setup, and the size-weighted hit ratio guide covers the second axis in depth.
Because a client that sends three hundred inquiries and trades ten is consuming three hundred prices for ten trades, and a client that sends forty and trades thirty is telling the desk it is competitive there. Flow alone puts the first client in the top tier; the hit ratio says what the flow costs.
Both are shown; the size-weighted one ranks. A client that trades the small inquiries and takes the large ones elsewhere has a high count-weighted hit ratio and a low size-weighted one, and the second is the truth.
Tier one gets the axe first and the salesperson's time. Tier two gets a pricing review, because the flow is there and the desk is losing it. Tier three gets more shown, because it trades what it sees. Tier four gets auto-priced. The tier is a coverage instruction, not a label.