Many businesses have become so enamored of the phrase "information age" that they end up buying too much information at too high a price. Such is the case with banks and brokerages, which routinely purchase a surfeit of market data on which to base trading and investment decisions.
Our analysis reveals, for example, that foreign exchange traders typically use only about 65% of the data their banks or brokerages buy for them.-rarely for less than $40,000 a year. That means $15,000 of waste per trader, or $20 million to $30 million a year for a top securities firm.
Institutions must pare these costs without, of course, alienating personnel and weakening the organization. The job falls to market-data management, which has to interact with information end users and external vendors to rationalize data demands and influence supply conditions and prices.
In some cases the major leverage resides on the supply side. More often, though, the demand side delivers the biggest economies and if properly managed can do so on a sustainable basis.
Proper management means institutionalizing a sound process that includes the following steps:
Don't let outside data vendors have unlimited access to the trader population. (No more free trials-they aren't free).
Inventory all available data.
Review end-user utilization of these data, using benchmarking techniques.
Having done all this, recast end-user entitlements to specific data feeds desk by desk.
In short, first determine what data you have, who uses which feeds, and whether they ought to be using them so often; then approve access only for those whose jobs legitimately require the information.
Assisting managers to curb excess demand is a piece of software that attaches to a digital "pipe" through which some vendor data reach traders' desks. The software, which allows market-data managers to entitle end users to the various data feeds, can be programmed to record usage rates by individual desk.
Presented with evidence of indifferent or sparing use, many traders will agree to dispense with the feed or accept a reasonable and cheaper alternative. Unfortunately, some market-data managers still fail to electronically compare data use with entitlement.
Sometimes, of course, traders can make frequent use of data feeds that they don't really need. How to establish need objectively? Obviously, through statistical analysis.
Which feeds do successful traders habitually use? In any given firm-and, in fact, on any given product desk-there is a large variance. Some traders of spot foreign exchange use as little as $10,000 to $15,000 worth of data, while others use $60,000 to $80,000.
Interestingly, the variance is just as wide among the most experienced traders as among novices. And stranger still, there is no direct correlation between data use and success. indeed, there may even be an inverse correlation.
Using this type of analysis, banks and brokerages can establish benchmarks or profiles for the data needs of traders of all types, which can then be reflected in a revised schedule of end-user entitlements. As a result, many traders hitherto addicted to browsing might encounter "data access denied" limits, but there is no evidence that these restrictions will adversely affect their earning capacity.
Having faced down the data junkies in their own shops, managers must turn their attention to two additional tasks: changing the data-technology infrastructure and renegotiating vendor arrangements.
The current technology infrastructure is seriously flawed, in large part because of the absence of uniform standards for both back-end data distribution and front-end data display. There are many platform types (for example, digital and stand-alone), and many vendors provide these platforms, each of which conforms to the vendor's proprietary standards.
This multiplicity of standards makes it cumbersome and costly to add one vendor's feeds to another's back-end distribution system-for example, a non-Reuters feed to the Reuters digital pipe. As a result, end users with the Reuters system are reluctant to switch to another data supplier for even a portion of their needs-a situation that is entirely to Reuters' liking.
Dealing with this situation requires market-data managers to encourage the development of common industry standards for distribution platforms.
The final stage of the cost-saving program involves head-on confrontation with data vendors.
The market-data industry is oligopolistic. The principal suppliers of data both in the United States and abroad are Reuters, Dow Jones Markets, Bloomberg, ILX, and Bridge. Most of these suppliers use their market power to maintain high prices and resist the unbundling of data inputs, preferring quite understandably to sell expensive complete packages even if the customer requires only a small portion of these packages.
But the market power of these vendors can be counterbalanced. That's because their customers are themselves potentially formidable oligopsonists-buyers who can organize to impose conditions on sellers.
At bottom, customer power in this industry stems from the fact that the buyers of information are also its creators: Their trades provide the raw material for the vendors. These gather the information from the exchanges or via direct contribution and then package and resell it to its originators: the brokerages and banks.
This suggests that unless the data vendors retain some semblance of reasonableness, the brokerages and banks can as a last resort organize to integrate forward, becoming not only atomistic suppliers of raw data but also organizers and disseminators thereof.
There is at least one example of a successful effort to achieve something very much like this. In 1993, to combat a threat posed by Reuters, 14 leading banks allied to form Electronic Brokering Service. EBS allows foreign exchange traders to view real-time bid and offer prices on their screens and execute trades electronically. It has since become the world's leading electronic forex broker, serving 25% of the London spot market in major currencies.
To be sure, successful efforts to discipline vendor pricing need not involve entry into the data-dissemination business. Significant price relief can be obtained by one or more of the following steps:
Negotiation with information-generating entities-for example, interdealer brokers-to provide data directly, thereby bypassing vendors.
Insistence on consistent global pricing schedules. (Vendors now charge more in countries where they face less competition.)
Use of size advantages to obtain volume discounts.
Insistence on the unbundling of data packages, to eliminate unneeded data feeds.
To summarize: reducing outsize market-data expenses is eminently feasible, provided you know what data you have; you know who is using which data feeds; you can establish by benchmarking who ought to be using which feeds; you can achieve some degree of standards uniformity; and you can stand up to the vendors.
None of these steps is especially easy, but the effort is well worth the trouble. Data-cost reductions of 25% to 30% have frequently been achieved.