BBI Contributing Editor

With all the vendor consolidation going on in the U.S. medical equipment industry, the number of vendors remaining in many segments is getting ever smaller. This raises questions about whether such mergers and consolidations are healthy for the industry overall or simply consolidate too much market share in too few vendors, leading to decreased price competition and slower adoption of innovative new technologies developed by smaller firms and startup ventures.

Each new merger of public companies is scrutinized by either the Federal Trade Commission (FTC) or the Department of Justice (DOJ). Companies not involved in this process may not understand how these groups can determine whether a specific deal is anticompetitive or not. While both groups use many things to make their final determination, one tool that is very useful is the Herfindahl-Hirshmann Index (HHI) or "Herfs."

The HHI calculation is made for the market before and after the merger. The difference indicates how much market concentration has occurred. The absolute value indicates the level of "concentration." The HHI is a simple calculation to make if you have the necessary data and it is accurate.

The HHI is simply the sum of the squares of each competitor's market share. If expressed as a percentage, it is multiplied by 10,000 to make it a whole number. For example, if there is only one company in a market, with 100% market share, the HHI for that market is 10,000 the maximum possible value.

If there were two vendors, each of whom had a 50% market share, the HHI would be: 502 + 502 = 5,000. If there were three companies one with 25%, one with 35% and one with 40% then the HHI would be 252 + 352 + 402 = 3,450.

If there were 10 competitors, each with 10% market share, the HHI would be = 1,000. So the smaller the HHI, the more competitors and more evenly their market shares are distributed. The higher the HHI, the fewer competitors there are or the more unevenly they are distributed.

Regulators set thresholds that describe the market based on their "Herfs." Markets with an HHI over 1,800 are "somewhat concentrated." Markets with HHIs over 2,600 are "concentrated." In concentrated markets, small gains in HHI that is, gains over 100 points can become significant to regulators who are reviewing proposed mergers.

The HHI gains are calculated by doing the HHIs before and after the merger and subtracting. For example, the Herfs for the pending GE Medical Systems (Waukesha, Wisconsin) acquisition of Instrumentarium (Helsinki, Finland), represented by Datex-Ohmeda (Madison, Wisconsin) and Spacelabs (Issaquah, Washington) in the U.S. hospital market, before the merger was 3,025. The HHI after the merger, calculated by adding the Datex-Ohmeda and Spacelabs market shares into GE's, would be 3,586, an increase of 562 points. Given that this is five times larger than the less than 100-point gains that normally attract regulators' attention in "concentrated" markets, it is reasonable to expect that regulators will take a closer look at this acquisition than if the gains were smaller (e.g., the companies being acquired were not such significant competitors with their own large market shares).

The regulators do not just look at the HHI for the overall market they tend to drill down into individual market segments. This is a good thing because it reveals not only where a problem may exist, but also how it may be resolved.

Segmenting the market varies from industry to industry. In automobiles, body styles sub-compact, compact, full-sized, van, truck, SUV, roadster are useful segmentations because they are both mutually exclusive and cover the majority of the market. An example of a less-helpful segmentation would be optional equipment things such as bucket seats vs. bench seats, AM vs. FM radios, optional lighting packages, standard vs. automatic transmissions. One of the reasons that these are not useful is that they are secondary to vehicle body styles. Any compact, subcompact, full-sized, SUV or van could have an AM or an FM radio, an automatic or manual transmission, etc. A customer shopping for a van with an automatic transmission, AM/FM stereo and bucket seats is unlikely to accept a roadster with the same optional equipment, because the body style van is the determining (or more important) characteristic. Regulators look for these first-level "determining characteristics" as segmentors in any market.

In the medical monitoring market, the first-level characteristic that is important is the clinical application of the product. If a hospital is looking for a surgical monitor (which requires identification and tracking of anesthetic agent and electrosurgical interference filtering) that is modular and stationary, it is unlikely to purchase a neonatal intensive care unit (ICU) or coronary care unit (CCU) monitor that is modular and stationary, but may not track anesthetic agents or have electrosurgical interference protection as these are the critical characteristics of the surgical monitor, not its modularity or configured design, its portability or stationary use. The hospital and surgeon or anesthesiologist may prefer a portable, configured monitor but this is important only within the group of surgical monitors that offer ESI filtering and anesthetic agent ID and tracking.

Going back to the GE Medical example or any other patient monitoring merger that has occurred or will occur in the future, the regulators will be interested in the clinical segmentations that each company has. Is there overlap in the CCU or the ICU or the ER or OR or PACU? Is there overlap in patient-worn telemetry ambulatory care areas or neonatal monitoring between the companies involved in the merger?

When one breaks the U.S. hospital market into such clinical applications and calculates the HHI gains for each clinical segment, the GE acquisition looks as it does in Table 4.

So as regulators look at the component segments of the U.S. patient monitoring market, they will see that the issues with the GE merger span virtually all market segments, with the possible exception of the emergency department, which seems to be a segment with more vendors and competition than other market segments.

With such HHI numbers, regulators would be inclined to take a second look at the merger, exploring ways that the deal could be approved that would not have such a dramatic negative impact on the overall market. One possibility in such situations is to spin off a component of the merger so that the market gains are not so dramatic.

In the case of the GE-Instrumentarium deal, it may be possible to take a piece out of the Instrumentarium "pie" and eliminate or minimize some of the issues. Since Instrumentarium has several component divisions, including Datex-Ohmeda and Spacelabs, a spin-off of the Spacelabs division, for instance, might have a positive impact on concerns about the HHI gains in the merger. In such a case, the "Herfs" would be recalculated again for the merger of GE and Instrumentarium (less Spacelabs) across all the same segments, and the impact reviewed. Our company, Medical Strategic Planning (Lincroft, New Jersey), has done just that in two new reports, "MSP 2310-Implications of the GE Acquisition of Instrumentarium" and the companion "Supplement Z-For Hedge Fund Managers." These reports are available from Medical Strategic Planning for $1,500 apiece, but readers of The BBI Newsletter can receive a $250 discount on each report.

In looking at such mergers, two types of market share data may be available revenue-based market shares and units-sold market shares. If both types are available, the more accurate to use would be units-sold market shares. These can be determined with much greater accuracy then revenue-based market shares. Indeed, units sold is one of the two components of revenue-based market shares. Revenue = units sold x unit cost. For an entire market, revenues = total units sold x average selling price. Total units sold can be determined independently of all the vendors. Revenues or average selling price cannot, as this data is buried in each company's income statement and can't be independently validated. Depending on the assumptions made for "average" selling price, you can have a very accurate knowledge of units sold and still get wildly inaccurate estimates of total revenues.

In many markets, the level of visibility of both of these factors is obscured, particularly at the individual market segment (clinical unit) level of the market. In our industry, however, the Medical Strategic Planning Reality Knowledgebase is available to track units sold across the entire market. This proprietary database tracks more than 400,000 monitors and defibrillators installed in more than 23,000 U.S. hospital clinical areas, as well as additional units installed in U.S. freestanding surgical centers and Canadian hospitals. It captures units sold to about 80% of all U.S. hospitals, a sample so large that extrapolation to the entire market can be made with considerable statistical accuracy. For this reason, we have used unit-sold vendor market shares in the above HHI calculations and not vendor revenue market shares.

There are other considerations besides the Herfs as well. For example, Instrumentarium acquired Spacelabs in July 2002, just six months before GE Medical moved to acquire Instrumentarium. This acquisition of a larger competitor by a vendor with a small U.S. market share elevated the combined entity into the middle tier of the market. What emerged was a new competitor that was larger and more viable than either of the two independent companies. This competitor was nearly as big as GE Medical and would have represented increased competition with GE, as it had products to sell in all the same markets as GE was selling, due to of the Spacelabs component of Instrumentarium. Since Instrumentarium's market share was small before the announced merger with its larger acquiree, there would also be a significant rise in the HHI due to this merger. However, in that case, there would be increased competition in the market, as the critical middle tier of the market which had grown very "thin" due to previous mergers and acquisitions, would be strengthened. GE and Philips Medical Systems (Andover, Massachusetts), both larger competitors, would have had to deal with a new and stronger mid-sized competitor. That is why Instrumentarium's acquisition of Spacelabs was approved quickly by regulators.

The situation for GE, however, is likely to be somewhat different. The combination of the second-largest vendor (GE Medical) with the just-emerged third-largest vendor (Instrumentarium) creates a larger second-place vendor (post-merger GE) and a market leader (Philips) who share more than 80% of the domestic market. But it leaves no real mid-sized competitors, making Datascope (Montvale, New Jersey), with its 6% market share, the No. 3 competitor in the sector, replacing Datex-Ohmeda, which had more than twice that market share.

From the regulators' perspective, this type of situation usually is not considered good for competition, as it concentrates too much market share in only two companies and disadvantages all of the smaller vendors who remain in the market after the merger.

So two different mergers, both with substantial HHI gains, can have opposite results, one increasing competition and the other reducing it. Obviously, HHI data by itself doesn't tell the whole story, but taken in concert with other information and common sense, the impact of a merger can be determined.

So the next time you hear of a merger in the industry, dust off your marketing plans and dig out your competitive vendor market share data and run the "Herfs" yourself. This will give you a good idea of what the potential impact on the market of the merger is likely to be.

(BBI contributing editor Arthur Gasch is president of Medical Strategic Planning, Lincroft, New Jersey. For more information or to order the market reports mentioned in this story, call 732-219-5090.)