Charlie Gibson, Chief Commercial Officer at Propellant Digital addresses how trading desks can optimise their use of CTP data
The fixed income industry has got the transparency it wanted with the new FCA and ESMA regimes.
The question now facing every desk that lobbied for increased transparency is: now that you have the data, how do you extract the most value from it?
Having access to a transparency feed and having a usable, decision-grade dataset are two different things. A sell-side algo desk and a buy-side portfolio modelling team are not asking for the same output, even from the same underlying trade reports. Data has to have been cleaned, normalised, validated and allocated to a specific job, to provide value.
It can then improve the whole investment process from research, backtesting, algo optimisation and portfolio stress testing through to pre-trade price formation. Whilst post-trade best execution, TCA, market share and performance analysis are all relevant.
As a former portfolio manager, I know how difficult it was to get a reliable picture of what had actually traded in a market; even something as (seemingly) straightforward as total volumes through the sterling credit market on a given day. The tape alone does not solve that problem, because you can never be fully confident that in-house cleansing has produced the right number. This is where the network effect becomes genuinely valuable: having clients across the market feeding back on the dataset, with those observations being acknowledged, acted on and validated by industry peers.
Where the data starts: Research and insight
On the sell side, research teams use a validated dataset to build market structure reports; monitoring trends in volumes across asset classes, tracking issuer activity and the most active ISINs, and informing new pricing strategies that account for demand across different curve points ahead of new supply. The most common use case is to feed a trading algorithm with a high quality, cleaned historic dataset, which is not only used to backtest strategies but also used to understand whether an enquiry was missed (in real-time) and, if so, where it printed.
On the buy side, risk and portfolio teams run the same high-quality historic dataset through liquidity scoring and average-daily-volume analysis at both the portfolio and individual ISIN level. This approach typically flags when liquidity in a name has quietly thinned out or become particularly active, which can then trigger a new order. The same dataset, viewed by peers, can be overlaid against internal positions to determine whether a new order is desired or indeed required.
Understanding the dataset and its nuances can also allow firms to build in early-warning triggers for risk mitigation. One example: ahead of the European bank default event, there was significant order flow over the preceding weekend. Had a firm been monitoring that activity, it could have built a trigger into its portfolio to generate an alert and potentially create orders based on pattern-matching against prior stress events.
Where the data travels: Pre-trade price formation
When a portfolio manager uses a liquidity event from the historic dataset to generate an order, that same data follows the trade to the trading desk. Feeding live prints into existing order management systems (OMSs) sharpens pre-trade transparency for any desk placing an order. Metrics such as average trade size, average daily volume (ADV), a position’s size relative to today’s market volumes, the last print price and size, and the protocol and venue on which the order traded all materially inform a trader’s approach to building or exiting a position.
What matters here is that a trader deciding where to route an order, or a risk team sizing a position ahead of a stress scenario are making that call on the back of post-trade data being read as a forward-looking signal.
“Get the underlying data wrong, and every decision built on top of it inherits that error. A single duplicated trade, for example, could double the apparent ADV suggesting a position that would take two days to exit could be exited in one.” Charlie Gibson.
Where the data proves its worth: Best execution and TCA
Once a trade is executed, the same dataset is asked to prove the outcome was a good one.
Best execution and transaction cost analysis (TCA) depend on a credible, like-for-like benchmark, built from a market view that is as complete and as clean as the desk’s own execution data. With a transparent, market-wide data set, European traders now have an objective benchmark for TCA.
Market share and flow analysis show where volumes actually sit relative to the wider market, rather than where internal records suggest they should sit. The same MiFID dataset also includes newer flags to identify on-venue matched principal activity and negotiated transactions, enabling firms to scrutinise their method of execution more precisely.
The data is increasingly used to validate pricing sources for product control and portfolio valuation; two functions that quietly underpin a great deal of downstream confidence, on the sell side and the buy side alike.
While the dataset remains the same, the question being asked of it, and by whom, changes. A consolidated, well-governed view of post-trade activity can answer a research question in the morning for a sell-side trading desk and a TCA question in the evening for a buy-side trading team.
Why confidence is the real product
As Vidal Mehra, Propellant’s Chief Product Officer, puts it: “The use cases only work if the data underneath them holds up to scrutiny. A best execution report or a stress test is only as good as the trade-level data feeding it, and that means duplicates must be resolved, deferrals correctly applied, and reference data aligned across regimes before a single chart gets drawn.”
Regulatory expertise is what allows a deferred trade under one regime to be reconciled against a real-time print under another. Technological expertise is what allows that reconciliation to happen at the volumes the new regimes now generate, rather than weeks after the desk needed the answer. Clean data needs network-level confirmation (not just internal validation) of the post-cleansed dataset, to assure firms that the market volumes they are working with match those seen by their peers across the market. That shared reference point is what turns a proposal from a mere test-case into something a risk committee, a compliance function or a client is willing to act on.
The industry got the transparency it asked for. The desks that benefit most from it will not be the ones with the most data, but the ones who trust what their data is telling them across every stage from research to the point of trade.
Transparency data touches the whole investment process, whether you operate with a short-term horizon on the sell side or a long-term one on the buy side. The key is having confidence in the measures derived from the raw data.
©Markets Media Europe 2025
