spend matters spend matters About this site
Advertise with Spend Matters
Advertise with Spend Matters
 

January 07, 2009

 

Treasury's $700 Billion Sourcing Challenge -- Is it Reverse Auction Time?

Quick, what's the best potential marketing plug for reverse auctions and related strategic sourcing approaches in the relatively young history of the strategic sourcing process? If you guessed the government's bailout of the failing mortgage and banking markets, you'd be right. According to an article in Monday's Wall Street Journal, "one likely method of purchasing and pricing assets is a reverse auction. In this, firms would offer to sell securities at given prices, and the treasury could buy the least expensive on offer. Institutions would presumably offer to sell at prices high enough to alleviate their woes but not so high they'd be passed over in favor of lower-priced offers."

A convincing argument, you might say, but is this a correct application of reverse auctions? Yes and no. I was trading emails this weekend with Sam Kinney (FreeMarket's co-founder and the process guru behind FreeMarket's auction models) about the applicability of reverse auctions in this circumstance. In my opinion, while a pure reverse auction model is ideal when an organization is attempting to drive to a market price based on highly specified parameters including but not limited to price and quantity, I believe a model that takes a hybrid optimization and reverse auction approach might be a better fit in the current circumstance. Here's why.

Unlike other markets where a transformed market leading price -- based on any number of factors from just unit cost to infinity – is the best solution, in the current case, it's impossible to gauge and weight all of the factors up-front. Indeed, Treasury is facing a moving target when it comes to buying up distressed debt and related obligations given that profit is only a single motive -- and one that's much further down the list -- in the scheme of things. Rather, preserving the integrity of the banking system and those institutions that participate in it is the primary goal. And this may mean it is in Treasury's best interest to pay more for certain obligations from certain participants than it is from others simply to keep certain organizations in business.



Or as The New York Times argued on Monday, "one of the most important decisions the secretary will make is the price the government pays for securities. Here again, there is wide discretion. He is directed to 'make such purchases at the lowest price' that is 'consistent' with the purposes of the 'act'. Those purposes, however, are expansive and leave him room to pay well over the lowest price available if he wishes to do so ... if he concludes that a higher price is needed to provide stability in the financial markets, that is evidently acceptable".

Sounds like a perfect time for a hybrid sourcing approach that combines the power of optimization based on running multiple scenarios and the competition of reverse auctions not to mention potentially reserving or opening up a following round after post-event analysis. How would it work? Treasury would provide feedback to participants based on how they stand under a variety of award scenarios on a constant basis over the length of the bidding event. That way, they could adjust their offers based on a variety of factors and see how it impacts their standing in real time. They might tweak bid quantities or respond with offers for similar securities but with slightly different maturities or expirations. The devil, obviously, will be in the details. But fortunately, there are a couple of off-the-shelf applications that Treasury could configure and customize to suit the purpose. This was not the case even a few years ago.

- Jason Busch

Comments
Jason- Thanks for the post and the mention. You are right about the complexities and the conflicting policy challenges involved. I have also wondered about a gain-sharing structure to these things. If Treasury pays $0.20 for something that eventually works out at $0.40, perhaps by some ratio the seller gets rebated some of the difference. This recognizes that there wasn't some magical business "value added" by the Treasury to raise the price from $0.20 to $0.40, but simply that Treasury "underpaid" on the day it bought for $0.20. Treasury in some ways becomes more of a pawn broker, holding collateral but the "seller" can retain the upside. This type of program would require CUSIP-number pricing, so Treasury knows the book value of everything it buys.

Adding complexity, I also learned this weekend that one of the investment banks, to "price" its CDO portfolio, used a model that ran for hours on over 4500 Sun servers. Evidence like this leads me to believe it may be harder to work fancy transformations and simply build a payoff matrix that hides this complexity.

Safe to say we're all learning minute-by-minute what this may require.
# Posted By Sam Kinney | 9/29/08 8:43 AM
Sam's got some additional thoughts quoted in a post over on a post on Supply Excellence:

http://www.supplyexcellence.com/blog/2008/09/29/ba...
# Posted By Jason Busch | 9/29/08 11:39 AM
# Posted By Alan Buxton | 10/1/08 2:23 AM
About Us | Advertising and Sponsorships | Advisory Services | Contact Us   © 2004-2008 Spend Matters Publishing, Inc. All rights reserved.