Thursday, February 17, 2011

What Watson’s Win Means For IBM’s Stock

Watson’s Jeopardy! win marks an amazing comeback for IBM, the 100 year old computer company that nearly perished in 1993. Back then, IBM was about to run out of cash before former McKinsey consultant Lou Gerstner took over and refocused the company on its core – helping businesses get value out of computing.  During its heyday, IBM was not necessarily the most innovative company, but it always shined when it came to looking at technology – not from a geek’s perspective – but on how it could help companies boost revenues or cut costs. 

Gerstner left IBM in the hands of Sam Palmisano and since then, IBM stock has risen – now standing at a near-record $163.40 with a market capitalization of $203 billion. Watson’s win highlights a technology that IBM expects to apply to helping financial services firms manage risk. And IBM is already using its analytics technology to help insurance companies boost the share of customers that renew their policies by finding better ways to serve them. Despite its success, IBM stock has further to rise based on its valuation and its recent track record of rising upside earnings surprises.

IBM developed a computer system that took 25 scientists four years to build at a cost of about $40 million, dubbed Watson, after its founder, Thomas Watson, that trounced the two best human Jeopardy! players in a two-day match televised Tuesday and Wednesday. With a two-game total of $77,147, Watson won more than three times as much as 74-game winner Ken Jennings -- who earned $24,000 -- and all-time Jeopardy money winner ($3.2 million) Brad Rutter.

Watson's technology has the potential to be much more than a publicity stunt. For Jeopardy!, it downloaded about 200 million pages of material -- including encyclopedias and movie scripts. When Watson got a question, say, in the category "Church and State", about Saturday Night Live's famous "Isn't that special!" comment by Dana Carvey, it searched those pages --known as its corpus -- and came up with hundreds of possible answers that fit some of the clues. Then, according to IBM researcher David Gondek, Watson applied an algorithm to assign a confidence number to each answer, based in part on the number of connections to the clue, and weighed those numbers -- picking the best one. In that case, Watson's guess -- "The Church Lady" -- was correct.

Business Value-Creating Innovation Drives IBM

To understand how Watson might help win customers, it helps to understand how IBM competes. The key to IBM's success was its ability to talk to a company's senior executives and explain how computing could help that company operate more efficiently or increase its revenues. The theme that propelled IBM forward was coming up with technologies that would enable it to sell such customer-focused innovation to businesses.

IBM has enjoyed an amazing comeback since its 1993 low point. According to Forbes, Gerstner started at IBM on April 1, 2003 and through a combination of cost cutting and changing the way it rewarded people -- from competing internally with other IBM groups to winning customers by providing competitively superior value -- he left about a decade later with the company in good shape.

Since then, Palmisano has carried forward Gerstner's decision to keep the pieces of IBM together. IBM's biggest business is services but its most profitable is software. According to IBM's third quarter 2010 financial report, IBM gets 57% of its revenues and 42% of its pretax profit from global services -- where it works with customers to analyze their business needs and develop technology solutions to them. It gets another 16% and 23% of its revenues from selling hardware and software respectively, and 0.3% and 44% of its pretax profits from those two. The balance of revenues and profits comes from IBM's finance unit. 

How Watson Could Boost Bank Profits

This strategy of customer-focused innovation has paid off for IBM, but how will that translate into putting Watson to work for business customers? According to IBM economist, Constantin Gurdgiev, one answer is to help banks and insurance companies manage risk. As Gurdgiev told me, global banks are lowering their targets for return on equity (ROE) -- net income divided by capital. Their net income is dropping because regulators demanded they exit profitable businesses like proprietary trading and their capital is rising because regulators are forcing them to put more capital aside to protect against risk.

Watson, Gurdgiev explains, can help such banks achieve their goals efficiently. For example, a bank could decide that it wants to boost its profit margins and Watson could pick the best way for it to raise the capital it needs to achieve that goal. Gurdgiev pointed out that Watson could develop many scenarios based on analysis of global capital markets -- including macroeconomic, financial, and strategic perspectives -- and possible global regulatory changes. Watson could then calculate odds for all these scenarios, pick the most likely one, and recommend a strategy for enabling the bank to achieve its profit margin goal.

IBM has already used a related technology, semantic analysis technology (SAT) -- that applies statistical analysis to identify ways to improve a business process -- to help insurance companies boost revenues. According to Christian Biecke, global insurance leader for the IBM's Institute for Business Value, SAT helped insurance clients boost insurance renewal revenue by 119% by enhancing customer satisfaction.

More specifically, SAT helped insurance companies identify dissatisfying elements of customer service and replace them with more personalized service that makes customers want to maintain their relationships with the insurance companies instead of canceling their policies. As a result, the customers renew their policies more often instead of bolting to a competitor.

Instead of simply looking at an insurer’s price relative to competitors’, IBM’s SAT examined customer psychographic factors. For example, IBM helped clients to recognize that a specific customer service agent, say, Joe Smith, seemed to be much more successful renewing policies from women living in southern California than those from Montana. And it found that another agent, say, Ann Morrow, was more successful renewing policies from the insurers’ female customers in Montana. As a result, IBM’s SAT recommended that its female California clients work with Smith and its female Montana clients with Morrow.

Does IBM's ability to use Watson and SAT to boost customers' profits mean you should buy its stock? With a Price/Earnings to Growth (PEG) ratio of 1.1, based on a P/E of 14.2 and earnings expected to grow 12.4% in 2011, IBM stock is fairly priced (since I think a PEG of 1.0 is about right). But the good news is that IBM has been providing investors with ever bigger positive surprises over the last few quarters -- most recently reporting earnings that grew 2.5% faster than expected.

As companies begin spending their $2.4 trillion cash hoards, investors could continue to be pleasantly surprised as some of that cash flows into IBM's business-value-creating coffers.


Post a Comment

<< Home