The Mainframe Blog Home

NASDAQ Needs a Mainframe

NASDAQ needs a mainframe with mainframe software engineering.

Oracle's Hardware Sales Down Sharply Again

Three months ago, Oracle reported financial results for the company's second financial quarter. Its hardware sales declined 14% year over year, to $953 million. Oracle predicted that hardware sales would fall again anywhere from 4% to 14% in the next quarter (at constant currency).

Actually they fell 16% year over year, to $869 million. Quarter to quarter they fell about 9%.

For perspective, Oracle claims that their nascent Exadata/Exalogic/Exalytics business is fast growing, but that only means the far larger Oracle/Sun Solaris business is collapsing even faster. Also, the earnings report reveals that "hardware systems support" (a.k.a. hardware maintenance) declined only 3% (constant currency, year over year). So Oracle's remaining customers are caught in the perfect storm of a cratering Oracle/Sun Solaris business combined with escalating maintenance prices. Fabulous.

Of course I realize that hardware sales can be cyclical, but Oracle's hardware problems are deeply structural. Oracle introduced servers with the new SPARC T4 processors in September, 2011, which took the SPARC CPU up to 3.0 GHz. This past quarter should have been a terrific one, or at least a decent one, given that Oracle/Sun model cycle. Instead it was awful again.

I predict that Stuart Alsop now has the opportunity to correctly predict when the last Oracle/Sun Solaris server will be unplugged.

My Mainframe-Related Pet Peeves

In no particular order:

1. "Green screens" are good enough. No, they're not. Do you force your users to submit their input and receive their output via punched cards? User interfaces change and evolve, and appearance often matters. Stanford's IBM mainframe served the world's first interactive Web application. If you haven't provided Web user interfaces on your mainframe to serve users' demands, what on earth are you waiting for?
2. Everyone must use Web interfaces. Some users prefer to continue with their familiar, fast, and efficient 3270 terminal user interfaces. Let them coexist. One size does not fit all.
3. We haven't implemented encryption yet. Every mainframe has built-in encryption support. Why are sensitive account numbers, Social Security numbers, credit card numbers, financial details, and passwords still flying around your network, internal or/and external, "in the clear"? Turn encryption on. Just do it.
4. FTP overuse. FTP is not an application integration solution! Connect two applications using FTP and you've automatically converted two or more business process steps into a "we might get around to it, eventually, if you're lucky" business process. Do you think your customers want that? And why are you copying all that sensitive data anyway? To make it easier for bad guys to get?
5. We don't allow TCP/IP connections to our mainframe "for security reasons." Congratulations, that "security" policy inevitably leads to the least possible secure environment you can imagine as the business finds every possible workaround to keep doing business — a true security nightmare. Let the z/OS Security Server and RACF do their jobs, please.
6. "Open" platforms and storage. If you connect exactly the same storage unit on your SAN (that you're already using for everything else) to a z/VSE system in exactly the same way, does that suddenly make your storage unit "closed"? If you're one of the people responsible for typing in activation keys to make sure Microsoft Windows can actually function, are you the same person who thinks that z/OS and Linux on z, both which eschew keys, are "closed"? Words should have consistent meanings. Many IT vendors have thoroughly debased the word "open," and some of us have fallen for that particular word game. It's past time they stop — and that all of us wise up.
7. "Mainframes are expensive." You know what's expensive? Not knowing the value of your financial holdings during a financial crisis because you've scattered bits of your portfolio records into little servers — that's expensive. Letting unreleased Michael Jackson records escape before you can monetize them. Billions of dollars of credit card fraud. Building yet another massive data center. Paying for 60 more licenses of Brand O middleware (this week). Adding another 20 staff to your payroll (this week) to support the IT mess you've implemented. You know what's not expensive? Mainframes. Stuff that works well isn't expensive.
8. "But that would require us to add MIPS...." So what? Business growth is never free, but it's darn inexpensive if it's a mainframe that's growing. And do you see MIPS listed as a currency, next to the yen, dollar, euro, and pound? It's not. IBM has different prices for different workloads.
9. Mainframe chargeback regimes. Everybody does them wrong. It's only a question of how wrong. Just because a mainframe, as a standard feature, lets you count and apportion various technical quantities like CPU-seconds doesn't mean they have much cost accounting significance. You certainly shouldn't be putting prices on those technical quantities while everything else in your data center (and beyond) remains uncounted, nor should those prices be different than true marginal costs (which can often be zero or near-zero).

Do you have any more I should add to the list?

IBM Announces 4Q2011 Earnings

Let's get right to the point: the mainframe had what's called a "tough compare" in the last quarter of 2011. A year ago, IBM's System z hardware revenues grew an impressive 69 percent year to year as the zEnterprise 196, with the world's fastest microprocessor, won tremendous favor among existing and new customers alike. That was probably the mainframe's best quarterly hardware revenue growth performance in history, at least when excluding the mathematically infinite growth rate in the long ago quarter when IBM shipped its first mainframe.

Every serious analyst forecasted that IBM's mainframe hardware revenues would be lower in 4Q2011 compared to 4Q2010, and that's what happened. However, there's an interesting and exciting twist. IBM had very nearly the same blockbluster quarter in terms of mainframe hardware capacity shipments that it did last year: MIPS shipments were down only marginally (4 percent). Moreover, the mainframe's gross profit margin was up. Profit is the mother's milk of R&D investment decisions for any corporation, especially a publicly traded one. And, with the application of simple mathematics, it's clear mainframe customers yet again enjoyed substantially lower per-MIPS pricing.

As I always point out when IBM announces earnings — and this time is no exception — mainframe hardware revenues are but a tiny part of IBM's, never mind the industry's, total mainframe-related revenues.

IBM's broader financial results are also quite interesting. While several other technology vendors — including Oracle, HP, and Microsoft — are reporting poor (or at least subpar) financial results, IBM didn't. So far IBM has proven to be remarkably resilient during economic downturns. In fact, there have been many times when IBM has turned recessions and depressions into competitive advantages. In one of the most famous historical cases, IBM's CEO Thomas Watson Sr. refused to reduce the size of his company or cut production during the Great Depression in the 1930s. IBM then was uniquely well positioned to win the new U.S. Social Security Administration's business to solve what was then the world's largest accounting problem. That business strategy was a huge, calculated gamble, but it certainly paid off for the company.

Anyway, this sort of earnings report is exactly what I like to see in a technology company. A company with stable, predictable growth in profits is extremely well positioned to make the long-term investments in research and development to assure a steady stream of exciting, useful technology products at competitive prices, including mainframes and mainframe software. Let's hope IBM keeps up its performance, but to date they're doing well.

Oracle Had a Rough Quarter

Oracle reported its earnings for its second financial quarter, which ended November 30^th. One of the big headlines: the company's total hardware revenues crashed below $1 billion to $953 million, down 14% year over year. Its software business was essentially flat on a revenue basis.

Oracle started shipping its new SPARC T4-based machines in late September. We would expect to see a substantial increase in revenue when a new model is introduced, but the opposite happened. Oracle's CFO explained that there were product transition problems which accounted for the drop. Perhaps, but Oracle is predicting that next quarter hardware revenues will drop again from 4% to 14% (constant currency). Keep in mind these figures include Exadata and Exalogic servers, which Oracle claims are growing. (CEO Larry Ellison said Oracle sold "over 200" of those servers combined in the quarter. More on that in a moment.) The bleeding continues unabated.

Another thing Oracle's CFO said should be disturbing to potential customers: "...we believe we could be back at pre-Sun operating margins shortly." If your revenues are declining, how do you increase operating margins? That's simple: you cut costs. Research and development are costs, and they should be big costs if you want to compete and win in the server market.

On the software side, Oracle has had some price increases, so a flattish performance there is also a problem.

Richard Sherlund, an analyst from Nomura Securities, asked a great question on Oracle's earnings conference call. In fact, Oracle promptly ended its call after his question. Sherlund asked, "Larry [Ellison, Oracle's CEO], could you reconcile these numbers you gave for Exadata and Exalogic? I think it adds up to 1,000 or 1,100. Didn't you say 3,000 just recently? And I think the company originally guided to 2,000." You can read what Oracle's executives said in reply.

In part of Mark Hurd's reply, he claimed "...we are taking huge amounts of share from IBM at their high end." IDC and Gartner keep close tabs on the server market, and they report exactly the opposite is happening. Maybe Hurd misspoke and will revise and extend his remarks.

European Commission Satisfied with IBM's Practices

As The Mainframe Blog and other outlets reported this past September, the European Commission and IBM reached agreement on a competitors' complaint concerning mainframe hardware maintenance services. IBM proposed changes in how it supplies mainframe spare parts and technical documentation. However, the European Commission wanted to hear comments from interested parties about IBM's proposed solution.

After that comment period, the European Commission is still satisfied.

In separate action, the European Commission is seeking more information on whether Google's proposed acquisition of Motorola Mobility runs afoul of European competition rules.

Japan's NTT Data Is Rock Solid with zEnterprise

NTT Data is the largest system integrator in Japan. In this video a couple of NTT Data's professionals discuss the new banking solution they're building for the Bank of Japan and the exceptional attributes of zEnterprise, z/OS, and WebSphere middleware products on z/OS.

An Update on the Relative Value of IBM v. Microsoft

As I mentioned this past September, IBM passed Microsoft in market capitalization. In other words, the financial markets assigned a greater total enterprise value to IBM than to Microsoft, based on the stock price multiplied by the total number of shares outstanding for each company. Stock prices are quite volatile, but it's interesting to see that the gap has widened. At the close of trading yesterday, Microsoft had a total market capitalization of $215.35 billion, while IBM's market capitalization was $228.71 billion. Translated into percentages, IBM is more than 6% more valuable than Microsoft.

Market capitalization represents the stock market's assessment of a publicly traded company's value, at least for the part of the company that's publicly traded. (Some companies are privately held with only a relatively small amount of stock publicly traded, but that's not the case here.) That value is, in turn, based on the market's assessment of the net present value of the company's earnings. The higher the expected profit (and profit growth), the higher the market capitalization should be, at least over the medium to long term. Therefore, at this moment, the stock market is betting that IBM has a brighter future in terms of profits than Microsoft, on a risk-adjusted basis.

What's particularly interesting to me in this comparison is that IBM's balance sheet, while extremely impressive, is rather different than Microsoft's. Microsoft has a bigger amount of cash (and cash equivalents) on its balance sheet. According to Yahoo! Finance Microsoft has over $42 billion in net cash, while IBM's total debt is actually greater than its cash balance. IBM's debt is a bit misleading because a good part of that is debt from its IT financing division, which is exactly what a financing division is supposed to do and which has no real equivalent at Microsoft. But there's practically zero valuation risk with cash on hand, meaning that the stock market should assign a lot of value to that cash. The stock market does, but it doesn't matter enough: IBM is the more valuable company. Said another way, the stock market doesn't seem to have much confidence that Microsoft knows how to leverage that cash hoard in profitable ways.

I'm compelled to agree, and the news this week reinforces my view. Consider that IBM is using some of its cash to buy Cúram Software. We don't know how much IBM is spending for Cúram, but a couple hundred million is a reasonable guess, perhaps even on the high side. Contrast that acquisition with Microsoft's $8.5 billion acquisition of Skype, which was completed less than two months ago. Is there anyone willing to argue that Microsoft's acquisition of Skype will be more than 42 times more beneficial (in terms of future profits) than IBM's acquisition of Cúram? I'm not willing to make that argument.

Meanwhile, Apple is still the world's most valuable publicly traded technology company with a market capitalization of $361.62 billion at yesterday's closing price. And with that statistic I'll conclude this blog post, written from my new MacBook Air, because my iPhone is ringing.

IBM Passes Microsoft in Market Capitalization

I mentioned a few months ago that it might happen, and it now has. The financial markets now assign more value to IBM than to Microsoft. That is, after the close of Wall Street trading on September 29, 2011, IBM had a slightly higher market capitalization than Microsoft. Market capitalization refers to the total value of a company's shares of stock.

It's hard to say exactly what caused this flip-flop in market positions. However, Amazon's announcement of their new Microsoft-free Kindle Fire, priced at only $199, might have nudged the market. Microsoft derives about 60% of its revenues from its twin Windows and Office software franchises, and, like the iPad, the Kindle isn't tied to either of those Microsoft products. Mobile devices are eroding (or at least containing) Microsoft's previously unassailable client device business. Those mobile devices also rely heavily on cloud delivery of computing services, almost always using Linux-based servers and middleware which Microsoft does not produce.

Microsoft fully capitalized on one of history's smartest business deals when, back in 1981, Microsoft agreed to supply the operating system for IBM's then-new personal computer. However, Microsoft insisted on a non-exclusive deal with IBM. Microsoft also insisted on retaining ownership and copyrights in their operating system. IBM was desperate for an operating system, particularly since Digital Research rebuffed IBM at least initially. IBM still very much viewed itself as a box-pushing company at that time, and IBM's lawyers and management agreed to Microsoft's terms. Later, IBM considered but rejected the idea of buying Microsoft (and Intel) outright. The rest, as they say, is history.

What's particularly ironic now is that arguably Microsoft relied far too long on a strikingly similar box-pushing business model (Windows and Office) while IBM, chastened in the early 1990s, figured out how to recreate itself, in large part recognizing the enormous value of having a rich, business solution-supporting software portfolio. That's not to say that Microsoft doesn't: its server-oriented middleware business is fairly large and one of the few moderate successes in recent years. However, Microsoft's middleware exclusively runs on Microsoft operating systems, and that's a challenge for the company and for prospective customers. In particular, Microsoft is losing a lot of potential business with Silicon Valley-based cloud-oriented companies, precisely the sort of customers Microsoft would have won in the past.

I think this market capitalization flip-flop helps demonstrate why the future of the mainframe is extremely bright. What every current or prospective mainframe customer wants to see is a stable, prosperous supplier with a great business model, growing sales, aggressive investment in the platform and its software to stay six steps ahead of competitors, continuous improvements in price-performance, and increasing architectural relevancy in areas such as cloud, security, etc. We're seeing all that, and that's great news for now and for the future.