Prediction Market for GSA Networx Enterprise: the Wisdom of Crowds Effect on the Probability of Telecom Bidding Outcomes

The question has been answered at last, Which of these Companies will be Among Those Awarded Part or All of the GSA Networx Enterprise Contract? The answer is, all of them. On Thursday, May 31st GSA awarded AT&T, Level 3, Qwest, Sprint Nextel, and Verizon availability to bid for domestic contracts worth up to $20 billion over the next 10 years.

Congratulations by the way go to Ulrich Tibaut Houzanme for being the "most correct" in his predictions with a total upside gain of $328.56 in the market we set up at Inkling - you're the proud new owner of the new Apple Video iPod we said we'd give away as the bounty for participating. For those of you who didn't get in on the action this time around, we'll be doing this again, so watch the newsletter for more details about future markets we're running. But first, a bit of analysis...

After the surprising announcement of the $48 billion Networx Universal award outcome in late March, this second portion of the U.S. Federal Government's telecom budget was particularly significant to all of the competitors for the future of their government sector business. So in the spirit of experimentation and since Jim Wisdom of Crowds Surowiecki was keynoting the SCIP Conference in New York in April, for the past 60 days, Aurora has been running a prediction market to determine probabilities surrounding the question, which proved to be strikingly accurate.

The stock of each company was equally distributed at $20.00 at market open, implying each company had identical odds at receiving a portion of the contract (five companies, 20 percent probability each). Almost immediately, the valuations of the stocks began to diverge, which we can suppose to reflect the aggregated information held by the traders in the form of the liquidity of their collective intelligence. That said, the depth of the market in terms of traders themselves was never more than low double digits, so the question of liquidity and probabilistic reliability is a valid one.

This initial volatility evened out over the next few weeks until, once the contract awards were announced on May 31st, the distribution of the stocks for each competitor hovered within one standard deviation of the opening price for each of them, with closing prices as depicted below:

• Sprint Nextel $25.69
• Level 3 $20.63
• Qwest $20.21
• AT&T $17.57
• Verizon $15.89

So what does this tell us about the Wisdom of Crowds phenomenon in predicting competitive outcomes? Well, there was no break-out leader of the pack, despite Sprint's clear probabilistic advantage, while Level 3 and Qwest barely moved from their original valuation in the end, and even the market losers AT&T and Verizon didn't experience dramatic price reductions that would suggest a more likely probability that they should be excluded from the award. In fact, the PM around this question was, probabilistically speaking, relatively successful. Had a single company emerged with a price range spiking from, for example, $20 to $65 or $75 while a majority of stocks in this market dropped to single digits, we could have probabilistically said a clearer winner was emerging in the mind of the market, which quite honestly, would have been far more satisfying for us as experimenters trying to predict an accurate outcome.

However, since each company ended their trading range relatively close to $20.00, let's dig into that differentiation of probabilities a bit. Why did the market predict Sprint Nextel at the top and Verizon at the bottom? Why was there no effective probability of a single winner, nor any of the competitors devalued to real long-shot status? Was the market predictive of the most probable outcome after all? To better understand the predictability of the PM and the outcomes achieved, the hopeful entrants must compare what the agencies needs are and what each company had to offer to meet those needs, while trying to understand the collective awareness of the market in aggregating that information across the depth of the trading population.

The Networx Enterprise contract articulated ten mandatory services to be eligible to be a part of the contract. These services are all fairly basic in the telecom services spectrum as compared to the 37 separate requirements on the Universal Contract. Did Sprint lose the Universal Contract because of some hidden perception that they could not satisfy this more substantial list of requirements or for some other reason, something as simple as comparative pricing? Either way, Sprint decided it wasn't necessary to protest the earlier Networx Universal decision, which in the spirit of good sportsmanship may not have hurt matters in securing part of the Enterprise contract along with providing the mandatory requirements. Whether that will ultimately matter much is debatable:

Because agencies have to choose whether to use Enterprise or Universal for each specific task, Sprint and Level 3 are at a disadvantage because they are not on Universal, said Warren Suss, president of Suss Consulting. Agencies are going to be balancing pricing and the range of services and range of locations that each contract offers, he said. My bet at this point is that most agencies are going to favor the Universal vehicle.

The Big Three theory of competition, says that even as AT&T and Verizon have solidified their hold on the number one and two spots, number three can no longer be thought of by default as Sprint, despite their scale and revenue numbers. Qwest appears to be in hot pursuit and is looking for just this kind of opening to recover from the woes the company has faced since the late 1990's days of largesse and scandal.

As the company perceived to have superior IP/Wireless services in the U.S., Sprint could conceivably be very competitive for a majority of the Enterprise contract. However, Qwest, who until the March announcement was a distant number four emerged from their Networx pursuit a much more confident competitor. Interestingly enough, Qwest's press release quotes were selected with care, such as this one from Mark Winther, vice president of worldwide telecommunications at industry analyst firm IDC, when he mentions that the, companies that win both the Universal and Enterprise portions of Networx are clearly the elite in the communications industry can be construed as intentional of Qwest's desire to compete on different footing than that which larger oligarchs have become accustomed.