What Oshkosh's FMTV win says about competition, industry structure, and intellectual property.

Last week, the US Army announced the winner of the recompetition of its supply agreement for medium trucks. The implications of the outcome merit some analysis.
Since 1992, the Army has purchased its Family of Medium Tactical Trucks (FMTV) from a factory in Sealy, Texas, about fifty miles west of Houston. (The Army refers to the entire family of related vehicles as the FMTV, but awkwardly, also calls any individual truck an FMTV. Someone please call William Safire about this.) The business was started by oilfield equipment firm Stewart & Stevenson; the original design was licensed from the Austrian truck firm Steyr, which is why American troops are found driving rather European-looking cab-over vehicles. Seventeen years and more than 54,000 trucks later, the factory is now owned by BAE Systems, and is the centerpiece of its Global Tactical Systems (er, Trucks) business.
At least, it was until about now. Years ago, the Army had bought the rights to the design from Stewart & Stevenson, and in May, bids were due in a build-to-print, fixed-price competition. Late on Wednesday, Oshkosh revealed that it had won the next contract from the Army, and that production (and at least 2,000 jobs) would be going to Wisconsin. For both companies, this was a huge deal:

• Oshkosh, after expanding several times over in the past decade, had been losing money recently, as its fire, trash, and concrete truck businesses had turned down in the recession. By the end of day Thursday, Oshkosh’s share price was up 23%. Moreover, after production of the FMTV winds down at BAE next year, Oshkosh will be the sole source of all medium (FMTV, MTVR) and heavy (HEMTT, LVSR) trucks for the US Army and Marine Corps.

• BAE Systems, having run the plant solely to produce the FMTV and its variants (LTAS, Caiman MRAP), may very well be left with nothing to do in Sealy after 2010. The company had hoped to win the MRAP All-Terrain Vehicle (M-ATV) contract, but Oshkosh landed that last month. Now, unless the Joint Light Tactical Vehicle (JLTV) program both proceeds and goes to the Lockheed Martin-BAE Systems team, the company will have nothing foreseen for filling the factory.

Indeed, as huge a deal as this is, we’ve gotten several questions from around the industry about it:

• What happened? We can’t say for sure, but it’s quite possible that Oshkosh simply beat BAE Systems on price. In the M-ATV competition, Oshkosh won with a bid of roughly $385K per truck; speculation before the announcement widely held that an M-ATV would cost about $500K. That suggests that Oshkosh just knows how to build trucks really efficiently, which is what one would expect for a company in a relatively competitive business.
• Does this mean that the Pentagon doesn’t care about a “competitive industrial base”? It’s just not relevant here. Colleagues have asserted to us recently that the Army Department, for one, has never paid much attention to industry structure. But even if it would, this wouldn’t be the place to start. In supply agreements for long-term contracts on commercially analogous items like trucks, there is little problem in putting as much business with one supplier as any customer cares. Allocating contracts on justifications other than quality and price is unhelpful here. We should all remember that Stewart & Stevenson entered the truck market with this (licensed) product back in the early 1990s. Thus, it's not just that the Army could turn to Daimler or MAN for trucks—market entry has already proven quite possible. There's no reason to agonize over future competition: wave enough money, and they will come.
• Could BAE Systems successfully protest the decision? Doubtfully. This competition was a build-to-print, fixed-price deal. Neither BAE Systems nor any potential supporters in the government could plausibly argue that Oshkosh couldn't be trusted to build trucks. Conversely, if the Army had given the contract to BAE Systems, for some sentimental reason, and yet at a higher price, that decision could not have survived a protest from Oshkosh The GAO would right now be ripping everyone's faces off.
• Was this all about the Army’s ownership of the technical data package (TDP)? That was necessary, of course, but not sufficient. The Army couldn’t buy Strykers from anyone but General Dynamics because it doesn’t own the rights to the design. It could choose to buy FMTVs from Oshkosh because it owns the rights, and because it doesn’t think that building to that print is too difficult. The Army could theoretically choose to compete its next upgrade to the Abrams tank (the proposed M1E), but it isn’t likely to do so, because this flagship vehicle is a tad more complicated, and a bigger deal if something goes wrong.

To some extent, then, one could say that this was an easy decision with relatively narrow implications. The relatively commercial nature of the product meant that Oshkosh could be trusted to deliver, and the high procurement volumes expected meant that Oshkosh could bid low even with a long learning curve. Oshkosh knows how to build trucks, so figuring out how to make the FMTV line lean shouldn't take too long.
But that might not be the whole story. Consider that on Friday, the Army released a request for information (RFI) regarding the upgrade of more than 1,000 MRAPs to the Panther Medium Mine Protected Vehicle (MMPV) standard. The MMPV carries the engineers of the Army’s route clearance companies; as the program was getting started in a Pentagon business-as-usual, buy a few hundred trucks sort of way, the MRAP program raced ahead to procure thousands. The Army now wants many more MMPVs for route clearance, but the Category II MRAPs lack some critical features, such as remotely operable bomb-poking arms and robot deployment ramps. Note further that the RG33L is the basis of the MMPV, and the Army has over 1,000 RG33 MRAPs. The RFI has indeed gone out to firms other than BAE Systems, signaling that the Army may trust third parties to undertake these scale and complexity of an upgrade, and on a front-line vehicle.
As one of our friends in the business put it,
all of this harkens to an idea that the several of us have been tossing about in various forms for a couple years. Competitive advantage in US defense contracting increasingly arises from anticipating customer requirements and providing them a pretty-good solution fast, as opposed to the traditional strategy of waiting for the customer to decide what he "requires", taking his money to develop a "solution" to the "requirement" at six percent margins, and then hoping for a long production run of the thing at ten percent margins. Contractors have always had the problem that the traditional approach buys them no sustainable advantage in intellectual property: they’d still have to fight the federal depots and other competitors for the aftermarket. The new twist on this paradigm is that even the production run is no longer a lock.
We’ve been alerted that more than a few Army program managers are, in the current budget environment, thinking about competing quite a few deals that would have gotten a pencil-whipped justification & approval just a short time ago. We think that rather leads to a to-do list for military product portfolios today:

• Think carefully about what rights you’re willing to sell to government customers, and for how much. In no case should rights be breezily signed away.
• Catalog extensively what intellectual property you’ve developed on your own, what you’ve developed with which customers’ money, and what you’ve jointly developed.
• When appropriate, hire some good patent attorneys and IP consultants to help you sort out the economics behind just what is worth how much.

There may even be hidden benefits to getting busy in this way. This is a technologically dynamic and engineering-intensive business. Taking full stock of what you’ve got both in the field and on the shelf could quite plausibly lead to commercialization opportunities that you never knew you had. And with declining military spending on the horizon (at least in the US), that’s a sound move to make.

Comments

[Jay C]

Do you think that OSK underbid the FMTV contract on purpose? I have heard that they were very aggressive with their bidding and hope to increase margins on the contract through engineering change proposals, similar to what BAE did. I don't exactly understand how these work but it seems that if the army (or OSK through a proposal) decides to change the structure of the vehicle for some reason down the road, OSK can charge extra for that at really attractive margins, since they already have the contract. They were banking on this.

Does this claim have any merit? If so, how much do you think they could have underbid the contract for?

[James Hasik]

Strictly speaking, I don't know that Oshkosh underbid. It certainly underbid BAE Systems, but whether it underbid its own costs (in the absence of anticipated change proposals) is yet to be proven. All the same, I'll have a bit more on this in a few days.

[bill keels]

there are two issues that i don't understand...

if i am correct, bae (formerly s&s) are using suppliers that they have developed, and those that are traditionally in the ohio/indiana/michigan automotive complex. Does transportation costs have much of an impact on cost work up? costs for transportation for heavy stuff (like used in these trucks) have skyrocketed in the rest of the manufacturing base Are any of the major components in the fmtv truck family GFE? such as engines?

second, the serious fraud office is reliably expected to issue a formal charge against BAE shortly in regards to alledged big big kickbacks in the al yamani omnibus contract between bae and the saudi government during the thatcher premiership. is there some hidden reluctance of us defense department to be potentially entangled in that mess?....and if bae is forced to make a huge payment to uk government to avoid prosecution, has that forecast been taken into account in bae's overhead burden on a new fmtv bid/contract....the numbers certainly look like it....the uk is certainly absent tooting the horn for bae on this protest as far as i can tell.

[bill keels]

i was by that plant in sealy a few days ago, and there must be 500-800 finished trucks in the yard....there is little activity taking place....NONE OF THE TRUCKS HAVE SERIAL NUMBERS STENCILED ON THE BUMPERS....have they not been sold off to the army for some reason, and still are the property of bae?...we used to see multiple convoys of low boy trailers with two each of the fmtv trucks going away from the factory in each direction on I-10..the factory is not on a railhead, so that didn't happen...is there a quality, spec conformity, or NFE (not fully equipped) issue going on?..usually you can see them running around the test track, but it is dead silent down there...

[James Hasik]

Bill,

You raise several interesting points. I'll respond as I can.

Transportation costs are definitely meaningful in automotive manufacturing, with such a high fraction of parts sourcing, and considerable weight to those parts. For that matter, the whole truck itself gets shipped from the factory to bases all over the US and around the world, so whether that factory is in Texas or Wisconsin, there will be freight to pay. If the program office is paying this, but not figuring the difference into the source selection, then the finished goods freight costs wouldn't matter. Shipping parts around would show up, so Oshkosh may be getting an economic edge from its location in Wisconsin. All the same, these costs wouldn't make up for the nine percent difference in price. After all, there are automotive factories all over North America, and that many of the vehicles sold here are built overseas, and people make money on that. So, transportation costs matter, but they're not decisive.

Next, while a great deal of the content in each FMTV comes from government-nominated suppliers, I don't know about major GFE items. There's clearly an economic difference if the program office doesn't take account of freight, regardless of who's paying, but I'm not sure. As for the engine specifically, it's a Caterpillar, but I *think* that it's just a directed item, and not GFE.

Like everyone, I've heard the rumors about the SFO's pending charges. I must say, though, that I doubt that the Army took this into consideration, or cares to do so. Without charges filed, there's not much that the source selection authority could legitimately do in that regard--or even should. Also, since this was a fixed-price contract, there's no reason to adjust BAE's overhead calculations. Neither a low bid on the FMTV nor a big fine would break the company, even if it cost BAE's shareholders a bit.

I do think, incidentally, that it would be markedly unfair to tag the culture at BAE L&A in the US with some of the charges that have been leveled about the ways of doing business at that particular aircraft unit in the UK. Whatever the merits of those charges, these are two very different organizations, with completely different corporate heritages. I'm not at all suggesting that as your intent. Rather, I'm just pointing out that the Army Department seems to consider the people at BAE, at least in the US, as an relatively upstanding lot. (Well, at least for us people in the defense industry.)

Lady Thatcher used to "go to bat for Britain" in military export sales, but I'm not surprised that Gordon Brown's government aren't going to bat on this one. First, at best, they'd be batting for Texas and some shareholders with UK passports. That's not likely to be a vote-winner vis-a-vis the Tories this spring. Besides, the MoD and the UKT&I Office have bigger fish to fry right now.

As for trucks possibly awaiting acceptance, that's hard to say. I don't know at what point the serials get stenciled, but hundreds of trucks awaiting shipment could suggest that the government found a systematic problem during post-production inspection. BAE and the program office publicly discussed a possibly similar FGI problem late last year, so this could be troubling. If the test track is not running, then it's possible that the quality issue can't be completely resolved without a test track run. That could be lots of things, but it's probably not a complaint about some unevenness in the clear coat.