Negotiation? Auction? A Deal Maker's Guide

The Idea in Brief

• What’s the best way to buy or sell an asset—through a negotiation or an auction? The answer depends on the nature of your buyers, the asset in question, and your own priorities.

• If you think you can get a sufficient number of the buyers you want to participate, an auction makes sense—unless you expect valuations to range widely.

• If you can write precise specifications for an asset, it’s probably wise to purchase it through an auction. But if the asset has the potential to create a lot of value, or a relationship or service is important, a negotiation usually works better.

• Also consider whether your deal requires secrecy, transparency, or speed. Auctions work best when transparency or speed is critical, and negotiations when discretion is a priority.

When you have something to sell, the best way to get a good price for it is to hold an auction, conventional wisdom tells us. The success of the online giant eBay is a monument to that belief, and a lot of academic research supports it as well. Auctions aren’t just for ordinary folks trying to dispose of antique teddy bears or 1970s wallpaper, either. Executives routinely auction off divisions or subsidiaries of their companies; after all, who can fault them for the prices they get if those prices are the outcome of a competitive bidding process?

Auctions have gained popularity with buyers, too, especially since the advent of the internet. Back in the early 1990s, many procurement managers began to see reverse online auctions, in which suppliers competed for customers’ business, as the way of the future. E-auctions, the managers believed, would allow them to quickly connect with a whole universe of potential vendors and reel in much more attractive offers.

In reality, however, auctions don’t always produce the most satisfying results.

Take a closer look at what happened with e-auctions. Many buyers soon learned that while online auctions were great at creating price competition, they couldn’t deliver meaningful competition on service and quality. At the same time, suppliers started to withdraw from e-auctions, believing they couldn’t earn reasonable profits by participating. But at the root of the pullback from e-auctions lay a more subtle problem: They set up a win-lose relationship between supplier and customer.

Reverse auctions can create the opposite of the collaborative supplier-customer partnerships held up as a best practice by many today. Forming such partnerships requires a more thoughtful process, one that allows buyers and sellers to exchange information and work toward an outcome that delivers the maximum value to both. In short, it requires a negotiation.

When to Auction, When to Negotiate

A number of factors—the universe of potential buyers, the nature of the asset, your own profile as the process setter, and the effects of openness—determine whether you should conduct an auction or a negotiation. Here’s a breakdown of which conditions are appropriate for each process.

Not all transactions require collaboration between the buyer and the seller, however. In many instances, an auction is still a better approach than a negotiation. The trick is in knowing which process to use when. To make that choice, you need to clearly understand your potential buyers, the characteristics of the asset in question, your own priorities, and the relative importance of speed and transparency to obtaining the best price. Let’s examine each of these factors.

The Nature of the Buyers

Sellers deciding whether to hold an auction should examine the pool of buyers from four perspectives: their number, their identity, their willingness to take part in an auction, and the valuations they’ll place on the asset up for sale.

Number of buyers.

The more possible buyers you have, the more likely you are to get a high price in an auction. But all else being equal, the value of each incremental bidder goes down as the total number of bidders goes up, as you can see in the exhibit “What’s the Optimal Number of Bidders?” At about 10 bidders, you’ll get 85% of the revenue that you could expect to get from an auction with 50 bidders. After about 15 bidders, the value that each potential buyer adds drops to almost nothing. In the absence of any particular advantage, each additional bidder has a smaller and smaller chance of being the highest bidder (who wins) or the second-highest bidder (who sets the price). And those are the only two bidders you care about in the end.

What’s the Optimal Number of Bidders?

After about 15 bidders, auction theory estimates, the value of having an extra bidder becomes negligible—as long as the bidders are similarly situated.

You might think that as long as the expected price keeps rising, you want as many bidders as possible to show up. But a wide-open process has downsides: First, there’s the complexity of managing so many bidders; and second, there’s the risk that some bidders (even high bidders) will choose not to play if the field is too large.

In many contexts, sellers consider the magic number of bidders to be somewhere between five and eight. When UK-based private equity powerhouse Apax Partners put Xerium Technologies on the block in 2002, it immediately winnowed 40 indications of interest down to seven bidders, on the basis of their reputation, credibility, and likelihood to close the deal. As Tom Gutierrez, the former CEO of Xerium, explained, “You can’t really handle more than seven serious bidders. It’s incredibly time-consuming.” It is possible to have too much of a good thing.

Certainty about who the buyers are.

If you’re confident that you know exactly who your potential buyers are, opening the process up to an auction might actually weaken your hand, especially if the universe of buyers is small. What happens if nobody else shows up? In that case, you would have been better off talking to the buyers directly. In other situations, you don’t know who the high-value buyers for your asset will be, and it would take a lot of time and effort to find out. In this scenario, search costs are high, so you should hold an auction, announce it to the world, and let the high-value buyers find you.

Buyers’ willingness to participate.

Are you sure the buyers you want will come to your auction? Many potential buyers may be wary because auctions can reveal participants’ identity and make their valuations transparent. They may worry that other bidders will get a free ride on their expertise. If specialized knowledge isn’t really needed to value an asset, you shouldn’t worry about deterring experts by holding an auction. But if such knowledge does matter, you may have to negotiate privately with the experts or at least hold a sealed-bid auction in order to bring them to the table.

Range of valuations.

Auction theory predicts that the final price in an auction will be the second-highest valuation plus a little bit more. This will be a pretty good deal for the seller if the top two valuations are close, because the winning bidder will come in almost at his or her price limit. But if the top two valuations are far apart, then the seller will leave money on the table by holding an auction.

Consider the auction held in January 2004 for the assets of Cable & Wireless America, which Savvis Communications won with a bid of $168 million. Savvis’s chief financial officer testified after the fact that operational synergies with CWA “may be as high as $60 million a year and may be higher,” implying that synergies alone would pay for the deal in three years. More important, before acquiring CWA, Savvis had been in a precarious position: It relied on two customers—Reuters and Moneyline Telerate—for 55% of its revenues. Such reliance makes financing difficult—and in the investment-heavy telecommunications business, access to financing is critical. Buying CWA gave Savvis hundreds of new customers and shrank the percentage of revenues that Reuters and Moneyline Telerate contributed to just 15%, reducing Savvis’s risk.

So CWA was valuable to Savvis in a way that was unique among the seven bidders present at the auction. The second-highest bidder, Gores Technology, was a well-known private equity firm. Although Gores had other portfolio companies in the same industry as CWA, buying it would have been primarily a financial play: Gores would have tried to improve CWA’s operations and capital structure and then sold the business for a profit. Unlike Savvis, Gores had no synergies with existing operations, no fixed costs that it could leverage across a larger customer base, and no benefits to realize from customer diversification.

Savvis’s stock jumped 33% on the announcement of the deal, increasing the company’s market capitalization by approximately $85 million. This implies that to Savvis, CWA was actually worth something in the ballpark of $250 million—far more than Savvis actually paid. CWA was a relative bargain, because Savvis had to beat the Gores bid by only a little bit.

The Attributes of the Asset

When choosing between an auction and a negotiation, buyers should take into account three aspects of the asset they intend to purchase: whether they can create precise specifications for it, the potential for value creation, and whether service or relationships are an important element of the deal.

Specifiability.

All else being equal, the more specific you can be about what you want to buy, the more likely it is that an auction will be your best approach. The ability to provide detailed specifications for an asset is sometimes, though not always, determined by whether it is a commodity. If you’re trying to buy paper clips in bulk, by all means hold an e-auction and let competition among suppliers drive down the price.

It’s very hard to capture the value of more-complex products and services in specifications. To begin with, many product specs cannot be determined in isolation from price. If product features themselves are up for negotiation or the options available aren’t well defined, writing auction specs will prove impossible. (For examples of how specifiability should affect the choice of whether to auction or negotiate, see the sidebar “A Tale of Two TARP Auctions.”)

A Tale of Two TARP Auctions

When the U.S. government established the Troubled Asset Relief Program (TARP), in October 2008, the initial idea was to hold reverse auctions to buy struggling banks’ “toxic” assets. The Treasury Department would specify a class of security and then banks would compete to sell assets that fit that specification to the government. The price would be bid down and the lowest price would win, thereby providing the biggest bang for the American taxpayers’ buck.

The problem was that there are at least 100,000 kinds of mortgage-related securities, with different vintages, maturities, interest rates, geographic locations, payment histories, and FICO scores. Even if the government specified the asset for each auction on the basis of just the most important features, there would be tension between creating precise enough specifications and getting a reasonable number of bidders to show up. In part for this reason, the Treasury ditched the plan a little more than a month after TARP had been established.

Fast-forward nine months. Several TARP recipients had repaid their TARP funds, which triggered their right to buy back, at “fair market value,” the warrants they had to issue to the government in order to receive bailout money. And how should fair market value be determined, given that there was no active market for these warrants? Goldman Sachs, Morgan Stanley, and other banks negotiated privately with the Treasury Department, fueling public concerns about backroom deals in which the taxpayers would be the loser. Jamie Dimon, CEO of JPMorgan Chase, decided to head off this criticism with a request that the Treasury Department auction off the TARP warrants that it held in his company. Unlike the toxic assets, the JPMorgan Chase warrants were entirely homogeneous and therefore amenable to auction. A transparent auction would also eliminate any concern about backroom deal making.

And so in October 2008 the U.S. government proposed to auction when auctions were not appropriate; and in 2009 the U.S. government chose private negotiations when auctions were quite appropriate. The two vignettes illustrate the importance of careful deal process design.

Potential for value creation.

The specification process itself can actually end up destroying some of the value you hope to gain from your purchase. Consider the following excerpt from a recent request for proposal (RFP) for a business service:

This description is not exactly conducive to a creative collaboration between the customer and the supplier, but it is just the sort of exercise managers get into when they attempt to prepare complex products and services for bidding. It also drives home the point that although auctions do a great job of boiling everything down to price, boiling everything down to price is often at odds with identifying valuable opportunities.

Imagine you prefer to have a project done by the end of the year but would be willing to accept a longer timeline in exchange for a lower price. Specifying a completion date in the RFP might deter bidders who could do the project for much less money if they had more time. So if there is some chance that a deal between a buyer and a seller can create extra value, it’s better to negotiate than to hold an auction. A negotiation allows a buyer and a seller to learn each other’s preferences, make trade-offs across different issues, and craft a deal that yields a larger pie for all.

Importance of relationship.

If a relationship or service is a crucial attribute of an asset, you should again lean toward a negotiation. A major reason that both buyers and suppliers have moved away from e-auctions in procurement is that an auction signals the buyer’s indifference to suppliers beyond the prices they quote. Procurement officers have come to learn that such signals can have negative consequences, especially when problems arise down the road.

For an illustration outside corporate procurement, consider the experiences of Edward, a Boston-area home owner looking to renovate a bathroom in his house. Many home owners would conduct an implicit auction among general contractors, letting each bidder know about the others as a way of ensuring a fair price. Instead, Edward picked Dave solely on the basis of word of mouth. In their initial meeting, Edward told Dave, “I’ve heard great things about your work, and I’m really looking forward to working with you. For this reason I’m not shopping around. But in exchange I hope and expect that you’ll give me a fair price and that we’ll work together to see this project through to a favorable completion.”

Sure enough, the project was done on time and on budget. But just as Dave was putting the finishing touches on the bathroom, disaster struck. Dave’s plumber came to the house to do some work in the basement and turned the water off. Soon thereafter, Edward’s cleaning person came to the house, opened the bathroom faucet, discovered no water running, and inadvertently left the faucet open while she cleaned the rest of the house. The plumber finished his work, turned the water back on, and left. When Edward returned home a couple of hours later, he discovered a foot of water in his newly renovated second-floor bathroom and water damage throughout the rest of the house.

Distraught, Edward called Dave and explained the situation. Both parties implicitly understood the legal backdrop: A court would have to apportion fault between Edward’s cleaning person and Dave’s plumber. But rather than engage in a complex and indeterminate legal inquiry, Dave repaired the damage and finished the project at no extra charge. In doing so, he remembered his initial encounter with Edward and the trust that Edward had expressed by not shopping around. Would Dave have done the same thing if Edward had held an auction for the initial business? There is no definitive answer, but my conversations with Dave afterward suggest that it would have been less likely.

The Process Setter’s Profile

Another major factor to consider in determining whether to auction or negotiate is your own profile. Two elements here are typically in tension with each other: speed and risk. Because an auction can happen far more quickly than a negotiation, it’s a better mechanism when speed is critical. But faster processes are also riskier, because you have less time to adjust to new information along the way.

Importance of speed.

Most of the steps in the negotiation process—such as exploring interests, generating options, and identifying ways to create value—take time. Moreover, negotiations tend to happen sequentially, because you can’t negotiate with two counterparties at exactly the same time. In some situations, you don’t have the luxury of time—there is a fixed window of opportunity, or the asset is deteriorating in your hands. In these cases, auctions dispense with the benefits of value creation in exchange for a quick sale.

Bankruptcy law implicitly recognizes this point. In the United States, “363 sales” (named after the provision of the bankruptcy code that permits them) allow companies to sell assets very quickly through auctionlike mechanisms. Section 363 is often referred to as a “rotting fish” provision, which comes from the notion that a bankrupt fish merchant is best served by immediately converting his inventory into cash rather than letting it rot while lengthy Chapter 11 proceedings unfold.

Tolerance for risk.

With speed comes more risk: You could end up with a “busted” auction, in which nobody shows up, or even worse an auction with only one bidder. In contrast, a negotiation lets you test the waters and carefully make your way to a better outcome.

Consider the case of a privately held dot-com that was contemplating a buyout of its minority shareholders in 2008. Another Harvard Business School professor and I were asked to advise the company on how to execute this so-called freeze-out transaction. One proposal was to hold a reverse auction, in which minority shareholders would name the price at which they would be willing to sell, and the company would buy back the shares from the lowest price to the highest until its buyback funds ran out.

We advised against this approach because it was too risky. There was some chance that minority shareholders wouldn’t be willing to sell at any reasonable price, preferring to hang onto stock in the hope that the company’s valuation would return to the soaring levels of the early 2000s. The dot-com could end up with a busted auction or an extremely unattractive clearing price. Even if the minority shareholders behaved more rationally, they still would be trying to figure out the bids of others while making their own bids, which meant that they might not truthfully disclose their willingness to sell. The results of a reverse auction were difficult to predict with any certainty and potentially disastrous.

The company decided instead to negotiate privately with its shareholders. The management team quickly homed in on two large minority shareholders—call them John and Fred. John came to the table first and offered to sell his stake for $12 a share, a sky-high number based on a valuation of the company at the peak of the internet bubble. The company countered with $1.50 a share, which had been the offer price in the first round of financing back in the late 1990s. It was a big gap, but the counteroffer had the effect of defusing John’s unrealistic expectations. The company came up a bit, to $3; John came down significantly, to $3.31; and the company quickly closed the deal at $3.31.

The negotiation also paved the way for a successful result with Fred. Fred was on the dot-com’s board of directors, and because of his fiduciary duties, felt compelled to approve this great deal for the company. But in so doing he had to acknowledge that the $3.31 price had set a powerful, and unfavorable, precedent for his own negotiation. In short, the company’s choice to negotiate rather than auction minimized its risks and allowed the dot-com to work its way to a very good deal.

The Effects of Openness

Finally, when determining whether to hold an auction or negotiate, you should look at two contextual factors: the need for secrecy and the need for transparency.

Need for secrecy.

Broad-based processes like auctions are difficult to keep secret. Even if you require bidders to sign confidentiality agreements, the chance of a leak increases with the number of people in the process. So if secrecy is important, you should consider negotiating privately with one or more buyers. Concerns about confidentiality often come up in M&A deals. If it becomes known that the company is for sale, employees start looking for new jobs, and the value of the company deteriorates while it is on the block. In my research I regularly find boards that resist a broader-based search for acquirers, because of this very reason.

The sale of the Pittsburgh Steelers, one of the most legendary franchises in the history of the National Football League, illustrates how this risk can play out. In 2007 one of the team’s co-owners, Dan Rooney, quietly offered to buy out the shares that his four younger brothers and their children held in the team. The terms of the offer weren’t publicly known, but whatever they were, the brothers weren’t satisfied. Working in tight secrecy under the code name Project Newcastle, they brought in Goldman Sachs to find a third-party buyer.

The story broke on July 8, 2008. The front page of the Wall Street Journal reported: “The Pittsburgh Steelers—owned for 75 years by the same local family—is secretly being shopped to potential buyers….” Not so secretly anymore.

For some assets secrecy is not terribly important. But for the Pittsburgh Steelers, the revelation caused an outpouring of negative fan sentiment. Despite the best efforts of Goldman Sachs, the broader search process that the younger brothers initiated quickly exposed the deal to public scrutiny. It’s possible that the negotiations with Dan Rooney had reached the point that the broader process was necessary to ensure that the younger brothers received the highest possible price. But the more open process also put the Rooneys on the defensive and tarnished an asset that had been a crown jewel of the Pittsburgh community for decades.

Unfortunately for the younger brothers, the global financial crisis broke in the middle of Goldman’s search. The one potential buyer turned up by the market canvass, billionaire Stanley Druckenmiller, walked away at the last minute, and Dan’s brothers came back to negotiate a deal. On November 15, 2008, Dan Rooney agreed to buy out his brothers in a deal that valued the team at $800 million—a number at the bottom of Goldman’s initial valuation range.

The younger brothers were clearly unlucky to sell the team in the middle of an economic meltdown, so we can’t fault the auction process alone for the price that they received. Still, it’s clear that their auction process increased the likelihood that the deal would become public and also signaled the younger brothers’ willingness to sell the team outside the family. Even though it was ultimately unsuccessful, this move damaged the franchise among devoted Steelers fans. Of course, in professional sports all is forgiven for those who win—and in January 2009 the Steelers won their sixth Super Bowl, setting a record among NFL teams. So in the end, the Rooneys’ flawed deal process may have been saved by the players’ spectacular on-field performance.

Importance of transparency.

The second contextual factor is the need for transparency, by which I mean the appearance of a level playing field for all bidders. When transparency is important, an auction is a better choice than a private negotiation. This is a major reason that most public procurement contracts and government-run privatizations involve auctions. They are particularly helpful when officials are looking to defuse accusations of corruption or favoritism—which may be why former U.S. Treasury Secretary Hank Paulson was attracted to the idea of holding reverse auctions to buy banks’ toxic assets in the fall of 2008. (Eventually, the challenges of specifying precisely the assets that would be auctioned scuttled the plan.)• • •

The world of deal making is not black-and-white, and choosing between negotiating the sale of your asset and holding an auction need not be an either-or situation. A deal process can very well combine a negotiation and an auction—a combination that I address in my forthcoming book. However, as you work toward a deal that enables both seller and buyer to fairly capture the most possible value, it helps to understand which factors tip the balance toward negotiation and which toward an auction. Structuring your inquiry in the way I’ve described in this article will help you come closer to the right solution.