Bid-offer spreads are traditionally explained by inventory costs, operating expenses and dealers’ risk of transacting with better-informed clients. In OTC (over-the-counter) markets, however, client knowledge and market power of dealers gives rise to price discrimination in favor of clients with high volumes and sophistication. Institutional investment strategies in forwards, swaps and options that are sensitive to transaction cost must consider the institution’s standing with their market makers.

Bjonnes, Geir, Neophytos Kathitziotis and Carol Osler (2016), “Bid-Ask Spreads in OTC Markets”, Brandeis University Working Paper Series, 2016-102, March 20, 2016.

This post ties in with a range of causes for the information-inefficiency of markets as summarized here.
The below are excerpts from the paper. Headings, links and cursive text have been added.

The traditional theory of bid-offer spreads

“Liquidity is arguably the most important product of financial markets, and the bid-ask spread is commonly the price of liquidity.”

“From classrooms to conferences to journal articles it is commonly stated that spreads reflect three specific dealing costs: operating costs, inventory costs, and adverse selection.”

  • Inventory costs are the market maker’s costs of holding securities of derivatives on his/her balance sheet. This includes capital cost and a premium for the risk carried. Inventories are a reflection of timing differences between supply and demand that occur even if the price ultimately clears the market.
  • Order-processing costs are the costs for executing a transaction. They are both fixed and variable. They largely relate to personnel and equipment but can be affected by the cost of search for counterparties, due to netting or other settlement economies.
  • Adverse selection costs are expected mark-to-market losses that result from trading with well-informed counterparties. They reflect the familiar phenomenon of ‘information asymmetry’. Traditional financial market theory assumes that dealers will charge a premium for transacting with clients that are often right about market trends.

Why OTC markets are different

“In auction markets the liquidity provider moves first by providing quotes which potential counterparties can view and compare directly; in traditional OTC markets the liquidity demander moves first by contacting the dealer, so a potential counterparty can see just one set of quotes at a time. If these quotes do not seem attractive the costs and risks involved in searching for a better price will sometimes dissuade traders from doing so. OTC dealers therefore hold a degree of market power at the moment they provide quotes.”

“Auction and OTC markets also differ in terms of pre-trade anonymity. When providing quotes in an auction setting a market maker cannot know the identity of his ultimate counterparty; in an OTC setting the market maker necessarily knows the counterparty’s identity. A rational OTC dealer will exploit this customer knowledge by price discriminating.”

“Price discrimination in OTC markets will reflect three customer properties.

  1. The customer’s market sophistication, meaning his familiarity with market conventions including customary mark-ups as well as knowledge of current market conditions. More sophisticated customers have greater negotiating leverage vis-à-vis dealers and are therefore likely to pay narrower spreads.
  2. The customer’s normal trading volume. Dealers will rationally seek to attract more active customers by providing volume discounts.
  3. The customer’s tendency to be informed about upcoming returns [information asymmetry].”

“These forms of price discrimination could help explain a consistent puzzling finding in the literature on OTC markets: a negative relation between spreads and trade size. This pattern has been identified for every OTC market tested in the literature…An inverse relation between information and spreads could therefore emerge naturally if the dealers’ tendency to price discriminate in favor of sophisticated and active customers outweighs any price discrimination on the basis of information per se.”

“Rational dealers with access to an interbank market will seek to attract informed customers, anticipating a brief window of time for profitable trading before other dealers learn the customer’s information… OTC dealers will price discriminate in favor of informed customers whose trades…arrive early enough to enable profitable exploitation in subsequent interdealer trades.

“By symmetry, the dealers will price discriminate against informed customers whose trades arrive too late for profitable exploitation… Many hedge funds and some other customer banks engage in high-frequency trading, often to exploit mis-pricings that disappear very quickly such as violations of covered interest parity and triangular arbitrage. These mis-pricing are probably gone by the time our dealer can process the information from a high-frequency trade – that is, price discovery is essentially over before the dealer can react.”

Empirical evidence for client-based price discrimination

“We…[provide] empirical evidence from the largest OTC market in the world, the foreign exchange market. Our data, which comprise the complete trading record of a top-twenty dealing bank over three months in 2012, are outstanding insofar as they provide exact mark-ups, trading venues, and customer identities…The sample includes 257,421 transactions.”

“Every mark-up in our dataset is tailored to the individual customer. The vast majority (over 95%) are set by an automated quotation system rather than by the active intervention of a salesperson or an interdealer trader. Even so, the mark-up algorithm is designed and parameterized by the salesforce together with a dedicated e-commerce team.”

“In this OTC market the price-discrimination component of a customer’s mark-up over the interdealer price ranges from two-thirds to six times the combined effects of operating costs and inventories. The influence of price discrimination thus meets or exceeds, in relative magnitude, estimates of the adverse-selection component.”

“Our results confirm that forex customers are severely penalized for any lack of market sophistication and enjoy generous discounts if they trade in high volumes

  • A one-standard-deviation increase in a critical measure of market sophistication is estimated to decrease the cost of liquidity on a given trade by 2.4 pips [a pip is USD0.0001 in EURUSD and hence close to 1 basis point], which is many multiples of the average mark-up of 0.4 pips and similar in magnitude to the average customer bid-ask spread.
  • A one-standard-deviation increase in a customer’s average daily trading volume is estimated to decrease the cost of liquidity by 1.2 pips.

Together these two dimensions of price discrimination dominate the inverse relation between information and average spreads.”

“OTC price discrimination against less-sophisticated customers can have a substantial influence on bid-ask spreads…Small and medium-size enterprises’ mark-ups [in the EURUSD market] are 22.7 pips, on average, and hedge-fund mark-ups are just 0.03 pips, on average. [One pip = $0.0001/€; at current exchange rates this is roughly equivalent to one basis point].

Dealers…price discriminate against informed customers whose trades arrive too late for profitable exploitation… We test this hypothesis by examining the influence of high-frequency trading. The high-frequency trades of hedge funds and customer banks are often intended to exploit triangular arbitrage opportunities and other evanescent mis-pricings. Our dealers’ trades with high-frequency trading customers arrive late in the price discovery process, triggering adverse selection.”

SHARE
Previous articleThe term premium of interest rate swaps
Next articleThe China credit risk
Ralph Sueppel is founder and director of SRSV Ltd, a research company dedicated to socially responsible macro trading strategies. He has worked in economics and finance for almost 25 years for investment banks, the European Central Bank and leading hedge funds. At present he is head of research and quantitative strategies at Macrosynergy Partners.