Algorithmic pricing is the practice of calculating bids and offers for a traded product via an algorithm. The inputs to the algorithm can vary. In addition to other bids and offers available in the market, an algorithm may factor in the depth of liquidity available at those prices (to anticipate the impact of a new trade on the market), and the market-maker’s own positions (adjusting the price one way or another to add or cut exposure). Rules or strategies will inform the algorithm’s behavior: a momentum strategy will assume market moves are, to a point, self-sustaining and will price accordingly; a mean-reversion strategy will assume the market self-corrects. Pricing algorithms are most effective when they have a lot of data to work with. Accordingly, they are most common in the most active markets – for example, equities, foreign exchange and government bonds. But their cost-cutting potential has also seen market-makers expanding their use into less-liquid products including corporate bonds and interest rate swaps. In these cases, the algorithm may suggest a price, which is then authorized or quickly nudged wider or tighter by a human trader; the same can be true for larger trades in liquid products.
ALGORITHMIC PRICING
