Over the past five years, there have been massive disruptions in the technology industry with established players ceding ground to new firms. A massive explosion in the cloud has rendered legacy hardware businesses far less profitable. Declining hardware sales also hurt follow-on service revenue, which is typically higher margin. In response, existing tech companies have worked to improve their cloud infrastructure to better compete with Salesforce.com (CRM) and others. While firms like Oracle (ORCL) and Microsoft (MSFT) have built fast-growing cloud divisions, legacy units continue to drag.
In the past year, two companies have been singled out as legacy tech bellwethers struggling to adapt to a shifting landscape: Cisco Systems (CSCO) and International Business Machines (IBM). As the following chart shows, shares have dramatically underperformed the broader market over the past year with worsening performance over the past six months. While these firms do face challenges, it isn't all bad news. As has been said many times, a CIO has never been fired for hiring IBM. Of these two struggling companies, only one makes sense in your portfolio. In this article, I will explore each company's operating performance, financial strength and valuation to explain why investors should buy CSCO and sell IBM.
(Chart from Google Finance)
For several years, IBM has been struggling with stagnating revenue, though a movement into higher margin businesses has helped to offset some of this problem. For instance, in the last quarter (available here), revenue dropped 5% and missed street estimates by $550 million, but operating income grew 8% while EPS of $6.13 actually bested estimates. However if you look closer at the results, investors will note that the company's effective tax fell to 11.2% from last year's 25.5%. Normalizing taxes for favorable one-time items, operating income would have been $1 billion lower or about 8% lower year on year, and EPS would have fallen 3% despite significant buyback activity.
Gross margins are relatively mixed at IBM. On a GAAP basis, they were down 0.1% to 51.7%, though they were up 0.3% on a non-GAAP basis. SG&A and R&D spending are also accounting for a larger percentage of revenue, and IBM has been cutting headcount to cut expenses. For all of 2013, revenue fell 4.6%. IBM is trying to push into cloud technology and services and is betting big on Watson. It is a cloud computing system that learns from new data and has been useful in diagnosing diseases at hospitals. IBM is trying to push Watson solutions throughout the business world. The company is hoping Watson can generate $10 billion in revenue within a decade, compared to the $100 million it currently generates (details on Watson revenue available here).
To reach $10 billion, IBM hopes to turn Watson essentially into a super-consultant to offer solutions to business problems, finance issues and so on, in addition to healthcare. Unfortunately, there are still problems that are stunting growth. Most glaringly, Watson isn't compatible with IBM's data center technology, which will make it far more difficult for IBM to regain its enterprise monopoly. Moreover, the company has been working from behind in the cloud revolution, leading to struggles in its server business. Other firms may also develop similar offerings, limiting Watson's upside and margins.
IBM is trying to figure out a way to grow Watson 100 fold over 10 years, which is a herculean objective. For a company that hasn't grown revenue in five years, that task may very well prove impossible. It also shows the challenge IBM's size poses. Even if it hit this target, Watson would only boost revenue by $10 billion in 2023 while the company generates over $100 billion today, meaning less than 1% annual growth. With current downward trends in IBM's revenue from existing business, IBM could very well have less revenue a decade from now than it does today even if Watson were to hit the target. To be truly successful, IBM will need to deliver three or four Watson sized successes. With less than $100 million in revenue, it looks like IBM could fail to deliver one.
Last in part due to the NSA scandal, IBM has been crushed in emerging markets. Growth markets fell 9% last quarter while BRIC nations were down 14% (data from the quarterly filing cited above). In some regards, these results were better than the third quarter when Chinese hardware sales fell around 40%. The NSA scandal should be fleeting, meaning sales will recover. However in nations like China and Russia, there is a push to be self-sufficient and use domestic tech companies rather than American ones. This will continue to put pressure on IBM's business in these growth markets. Overall, IBM's business faces serious challenges. Revenue is falling and margin pressure is starting to build in the legacy business. New-cloud businesses are helping to offset some of this decline, but IBM will need to fully deliver on Watson, which is unclear while developing several more groundbreaking technologies. Emerging markets should get better but will also be a growth headwind over the next 24 months.
Cisco faces the similar problem of legacy business weighing down results. However, its difficulties appear steeper as its businesses decline faster. In particular, set-top boxes have been disastrous with revenue down 22% and forward looking orders down 20% (full results available here). A turnaround in this unit is highly unlikely due to high competition and cord-cutting. This is a business that will decline all the way to zero. Cisco's traditional switch and routing businesses are also falling 10-12% due to cloud computing and a decline in land-line activity. New orders have shown a deceleration in the decline to about 5%, which is still relatively steep. Declining sales and increased commoditization have crushed gross margins with product gross margins essentially at a decade low of 48.7%. While IBM is struggling, Cisco's legacy businesses are performing even worse.
Now, Cisco has been building a better portfolio of new technology products. In particular, wireless, data centers and security are powering some growth. Overall, cloud related orders are up double digits. Security revenue has been growing mid-teens, and new orders are up 30%. The Meraki platform is doubling in size, which should push wireless revenue up over 10% going forward. Data center revenue has been increasing double digits, and new-orders are up over 30%. While these businesses are not yet big enough to fully offset the declines of legacy businesses, they are improving.
Cisco also has been hurt by the NSA, but its emerging market performance is improving and a bit better than IBM. International orders were down 3% last quarter, far better than the previous quarter's 12% drop (details in the quarterly conference call transcript). The BRICs are lagging at -10%, likely because China continues to avoid American firms whenever possible. In total, both firms face challenges but are developing new businesses to offset the decline. IBM's legacy businesses are declining more slowly than Cisco, though Cisco is performing a shade better overseas. In aggregate though, IBM's operating profile is slightly better than Cisco's, though Cisco looks like its revenue should stabilize towards the end of the year.
Cisco and IBM are also using their balance sheets and cash flow to finance buybacks and dividends. Cisco has upped its repurchases to a $4 billion quarterly pace and should fulfill its $15 billion (12% of shares outstanding) commitment this year. It also pays out a $0.19 quarterly dividend (up 12% year over year). While Cisco is starting to more aggressively return capital to shareholders, IBM has been doing so for years with a recent $15 billion authorization. Including the $5.6 billion left on a previous plan, IBM should repurchase 9-10% of the float over a twelve month period. It also pays out a dividend, though the yield (2.10%) is lower than Cisco's. IBM has been returning cash more consistently, but Cisco is starting to return more as a percentage of its market cap.
Cisco also has better financial flexibility to increase the pace of capital returns. IBM carries $11 billion in cash against $33 billion in debt. Considering its $200 billion market capitalization, its net debt load of $22 billion is not overly burdensome. Still, it does not have excess cash to return to shareholders. All capital returns will be from free cash flow, asset sales and debt issuance. IBM should generate about $15 billion in free cash flow this year while its dividend costs $4-$4.5 billion and share buyback could total $15-20 billion. In other words, IBM is already returning more than it is generating, so IBM's results will need to improve to sustain this pace of capital returns.
On the other hand, Cisco carries $47.1 billion in cash against $17.1 billion in debt for net cash of about $30 billion. In other words, Cisco has existing cash to spend and can afford to generate less free cash flow than it returns to reduce its overly large cash hoard. CSCO should generate over $11 billion in free cash flow this year, so it will eat about $6-8 billion from its cash hoard. Given its large cash position, Cisco can afford this drain for several years. On the whole, Cisco has a better balance sheet, which makes its capital return strategy more sustainable.
Finally, we reach valuation. Cisco should earn about $2.00 in 2014 and generate $11 billion in free cash. IBM should generate $14-15 billion in free cash flow and earn about $18 thanks to its repurchases. This means CSCO trades at 11.25x earnings while IBM trades at 10.7x. Cisco has a free cash flow yield of 9.17% while IBM has a free cash flow yield of 7.5%. However as mentioned above, Cisco carries a lot of cash (about $5.60 per share). If you cancelled out Cisco's cash, it trades 8.5x earnings and has a free cash flow yield of 12%. In other words, Cisco is substantially cheaper than IBM.
On the whole, Cisco has a slightly weaker operating profile than IBM, but its capital returns are more stable thanks to its superior balance sheet and has a cheaper valuation. On an earnings basis, Cisco shares trade at a 20.5% discount to IBM, and on a free cash flow basis, the discount is 36%. While Cisco has seen steeper declines in its legacy businesses, this discount to IBM is extremely exaggerated thanks to stronger growth businesses. Given its challenges, IBM is fairly valued or slightly overvalued at current prices. Cisco is appropriately valued at around 5-10% less than IBM, suggesting fair value of $25-$26. At current prices, investors would be wise to sell IBM and buy CSCO.