Will Rhode, TABB Group, US Swaps: The Future is Emerging, here.
Among the key action items we conclude that:
- Dealers need to continue competing for swaps clearing market share. Not only is clearing set to become a healthy revenue stream in its own right but we see an implicit decision by the buy side to send swaps execution flow to their clearing broker;
- In the face of shrinking ticket sizes and lower volumes, dealer revenues will come under pressure with the introduction of SEFs. Each bank will need to decide on how it chooses to implement market share strategies, whether it be via relationships and the phone, via SEFs and algorithmic pricing, or some hybrid of the two;
- The disintermediation effect of Dodd-Frank notwithstanding, dealer relationships with the buy side will continue to be key. The trick will be in knowing where the relationship really lies, across the array of potential business lines and myriad of individuals, and understanding why a buy side firm wants to engage.
Robert Almgren, Quantitative Brokers, Market Microstructure, Quantitative Trading and Interest Rate Markets, here.
Fixed income modeling is a well-established, extensive area of quantitative research. But most of that research focuses on pricing of the products themselves and their interrelationships via yield curves and the like. Here we want to point out some of the features specific to how these products trade in real markets that make them fertile subjects for ongoing quantitative research.
1. Futures products in general trade on single exchanges: competing products on different exchanges are not fungible. And the vast majority of trade activity is now electronic rather than in the pit; market data is extremely accurate and reliable. This is in contrast to equity data, where it can be a challenge to match trades against prevailing quotes.
2. Interest rate products are strongly affected by macroeconomic information announcements such as employment, inflation, etc., as well as by government bond auctions. Unlike equity earnings announcements, these information events happen in the middle of the trading day. To trade these products effectively, it is necessary to have a good understanding of which events are important, and how they will affect different markets. As an example, one finds that US information announcements have strong effects on US and European markets, but European announcements have very small effects on US markets.
3. Interest rates at different maturities are very strongly interrelated, much more than any pair of stocks. The markets are inherently multidimensional. Even if one is trading only a single contract, one must take account of the entire universe. Co-integration, a well-studied but rarely observed property of financial time series, is ubiquitous in this space.
4. Also because of the interrelationships, spread products and basis algorithms are extremely important. One may not have a belief about whether rates will go up or down, but one may be willing to bet on a change in the slope or curvature of the yield curve. Futures exchanges offer a variety of calendar spread contracts, as well as more complicated combinations. These spreads can give rise to “implied liquidity,” which is an important resource for trading. That is, an order in a particular maturity may be filled by a combination of different maturities and spreads.
5. Because short-term products generally have low volatility and large spreads, exchanges commonly use matching algorithms that are more complicated than time priority. For example, in pro rata matching, an incoming market order is matched against every resting limit order at the best price, independent of time of submission but in proportion to the size of the limit order. This greatly changes the optimal strategies for limit order placement, and the market dynamics following large trades. Some exchanges experiment with combinations of pro rata and time priority for products of medium duration, and the actual rules can become quite complex.
Order matching Algos, Algorithmic Trading – an Introduction, here.