Institutional OTC execution is often compared on the number that is easiest to see: the quoted spread. That is understandable, because a tighter quote appears to offer an immediate and measurable saving. Yet a quote is only useful if it is firm, available for the required size, accepted without friction and converted into a completed transaction close to the economics the institution expected. A desk that looks cheapest at the point of RFQ can therefore prove more expensive once rejections, delays, hedging exposure, settlement and capital usage are included. The mature way to compare OTC execution is to measure the whole path from request to usable capital, not just the first price on the screen.

The visible price is only one layer of cost

Traditional markets have spent years developing transaction cost analysis because headline pricing can hide the real economics of execution. As how institutional crypto investors are moving beyond surface metrics reflects a broader institutional shift toward measuring what execution actually costs beyond the quoted price. The quality of a trade cannot be inferred from the displayed spread alone. Crypto OTC is not identical to wholesale FX, but the same analytical instinct is useful when counterparties compete on apparently similar quotes. Institutions need to know what happens between requesting a price and completing the trade. That path contains costs that may never appear as an explicit fee.

Consider two desks quoting the same block. One shows a slightly tighter spread but regularly requires a second price check when the client accepts, while the other provides a marginally wider but genuinely firm price for the stated size. The first desk may win a screenshot comparison and lose an economic one if the market moves during that extra step. The difference is not theoretical, because a large order exists in a moving market while the execution protocol decides whether the quote will actually stand. A useful comparison therefore starts with firmness rather than tightness.

Firmness changes what a quote is worth

A firm quote transfers short-term price risk to the liquidity provider for the agreed acceptance window. A non-firm or conditional quote leaves more of that risk with the client, even if the displayed price initially looks better. Institutions should therefore track how often quoted prices are accepted as shown, how often they are refreshed, and how execution changes when volatility rises. Those metrics reveal whether a desk is consistently offering executable liquidity or merely attractive indicative pricing. The tighter quote is valuable only when the client can actually trade it.

This becomes more important as order size grows. A desk may be competitive for a modest RFQ and materially worse for a larger one if its available liquidity is shallow or fragmented. Asking for "the spread" without specifying size is therefore like asking for cloud latency without specifying the region, workload, or percentile. The answer can be technically correct and operationally useless. Institutional evaluation should compare price, size, and firmness together.

Implementation shortfall is the better question

The most useful analytical question is not "what did the desk quote?" but "how did the completed transaction compare with the decision price available when the institution chose to trade?" That gap is a form of implementation shortfall, a framework widely used in institutional transaction cost analysis. It captures the fact that execution quality includes both explicit and implicit costs, including market movement while the order is being completed. In crypto OTC trading, the framework can be adapted to compare RFQ response, acceptance, execution, and settlement timestamps rather than treating the quoted spread as the whole result. The objective is not to force every crypto trade into an equities formula, but to stop losing information between the initial decision and final execution.

This also helps separate market conditions from counterparty performance. If every desk becomes more expensive during a violent move, the market itself may explain much of the change. If one desk consistently shows more slippage between quote and completed execution than comparable counterparties in similar conditions, that is different information. A useful TCA process therefore normalizes for asset, size, volatility, time of day and liquidity regime before drawing conclusions. Without that context, institutions can mistake easy market conditions for strong execution.

Rejections and response time belong in the scorecard

Institutional FX provides a useful warning about focusing too narrowly on displayed price. The Global Foreign Exchange Committee has long required transparency around last-look practices, where a liquidity provider can accept or reject a trade request after receiving it, and has stressed that clients need information to evaluate how those requests are handled. During stressed markets, reject rates can rise sharply, which means a nominally attractive stream may become less valuable exactly when liquidity is most needed. Crypto OTC protocols differ across desks, but the procurement lesson travels well. Rejection frequency, requote frequency, and response time should sit beside spread in the execution scorecard.

Speed needs similar nuance. The fastest RFQ response is not automatically best if it is indicative, based on insufficient size, or likely to be refreshed before acceptance. At the same time, a slow response creates exposure because the market can move while the institution waits. The useful metric is therefore not raw response latency but the time from RFQ to firm, executable price at the required size. That turns speed from a marketing statistic into an execution-quality measure.

Settlement and prefunding are trading costs too

Once a trade is executed, the economic comparison is still not finished. Different counterparties can require different settlement assets, prefunding arrangements, banking routes, and operational steps, all of which affect how quickly capital becomes reusable. A quote that saves two basis points can be economically inferior if it requires significantly more idle capital or creates a longer exposure window before final settlement. For institutions trading repeatedly, capital efficiency compounds across transactions. Settlement should therefore be measured as part of total execution cost rather than treated as a back-office footnote.

The important point is not that faster settlement is always better at any price. Institutions may rationally accept slower settlement for stronger legal protections, a preferred custody model or a particular fiat rail. What matters is making the trade-off explicit, rather than letting it disappear behind the headline quote. The scorecard should record settlement time, prefunding requirement, asset and currency flexibility, failure rates, and the point at which the resulting funds become available for reuse. Only then can procurement teams compare like with like.

A better OTC scorecard is multidimensional

The practical output of this analysis is not a single universal formula. It is a consistent framework that forces counterparties to be compared on the same dimensions across enough transactions to reveal patterns. Quoted spread, effective spread, fill size, quote firmness, rejection rate, time to firm price, implementation shortfall, settlement time, and prefunding requirement each answer a different question. None should automatically dominate the others, because institutions have different priorities depending on asset, jurisdiction, urgency and trading frequency. The value comes from seeing the trade-offs rather than compressing them into a misleading league table.

This framework also changes the conversation with OTC providers. Instead of asking who can show the tightest BTC quote during a quiet hour, institutions can ask how execution behaves when size increases, volatility jumps, a banking rail is closed, or one source of liquidity disappears. They can request distributions rather than averages, because the tail of execution performance often matters more than the mean. They can separate quoted competitiveness from operational reliability and identify whether a desk's apparent advantage survives difficult conditions. That is a more demanding conversation, but it is also much closer to how institutional counterparties are evaluated in mature markets.

Best execution is a path, not a price

The spread remains important, and no serious institution should ignore it. The mistake is treating it as a complete measure of OTC quality when it captures only one moment in a longer process. A genuinely cheap trade is one whose quote remains executable, whose size is real, whose completion does not introduce avoidable slippage, and whose settlement does not consume more capital than the headline saving justified. Once those dimensions are measured together, the ranking of counterparties can change significantly. The tightest quote wins only when the rest of the trade confirms that it was actually the cheapest.