IT Investment Research

TaxAct, ERP and the End of the "Bush Tax Cuts"

Neither the imminent end of the 2001 and 2003 "Bush tax cuts" nor H&R Block's (HRB) recent acquisition of SecondStory, the makers of TaxAct personal productivity software, really mattered much to me when I first read about them a few weeks ago. 

On one hand I don't have to worry about the much maligned high end of the expiring tax-table income scale nor -- conversely -- the demise of the 10% bracket. I guess I do lose the 5% of the upper bound of the lower tax bracket but that's just a couple of hundred bucks, right? And I get hit with the marriage penalty again but life circumstances have changed since 2001 and it might make sense now for my wife and I to file separately. And if the rates change on the middle brackets a few points, "Hey, 2000-2009 was a good decade.  But all good things must end."

And although I do use TaxAct, when I first bought it I did not care that the company behind it was about 10 years old and founded by a couple of guys in Cedar Rapids so why should I care now if the software is now the responsibility of a big multinational accounting firm. I can still use next year's version of TaxAct to see if my wife and I should file separately.

But on October 28, the two seemingly unrelated news items crossed paths.  Bloomberg reported in a news story that was widely distributed in print and on the web that "middle class taxes" were going to increase 50% in 2011 if the Bush tax cuts were not extended.  Specifically the article said

"For a married couple with an income of $80,000, that would drain an extra $221.48 in withholding from a semi-monthly paycheck, according to calculations by the Tax Institute at H&R Block."

That's an extra $5000 a year to the IRS, not the 10% or so increase in taxes (about $1200) that I had estimated on the back of an envelope (and had already figured out how to defer or avoid). True the Bloomberg story -- based, as the above quote says, on H&R Block calculations -- is based on IRS withholding tables as opposed to IRS tax tables.  But certainly the U.S. government does not encourage people to withhold $3800 more than they are likely to owe does it?

So I googled around the web and found a source that says the tax increase likely to hit middle class families in 2011 due to the expiration of the Bush tax cuts is more in the lower range I had expected. Nick Kasprak at the Tax Foundation has put together an interactive, web-based tool to simulate its effects (see www.MyTaxBurden.org).  The Tax Foundation tool is more detailed and transparent than H&R Block's.

I don't know which source has the numbers right. But if H&R Block has it wrong, do I want to continue to use TaxAct?

And if experts apparently disagree on this little bit of basic personal tax calculating when it comes to simple income taxes, imagine the effects of large upcoming U.S. government regulatory changes [e.g., Obamacare, Frank-Dodd (or is it Dodd-Frank?), EPA-imposed cap and trade] on relevant ERP software. 

(As an aside, adapting enterprise software to such changes is why IT users pay what many IT analysts call onerous maintenance fees. Software maintenance is about content, not fixing things.)

-- Dennis Byron

(No financial interest in companies mentioned)

Ozzie's "New Day Dawning" Memo Does Not Say the End is Near for Windows

Microsoftie (or about to be former Microsoftie) Ray Ozzie is out with a blog post the week of October 25 disguised as a final memo to Microsoft (MSFT) staffers and peers. Or is it a memo disguised as a blog post?

Either way it deserves more than the predictable five-paragraph news stories that I am seeing that basically describe it as some kind of kiss-off of Windows. Of course it also deserves more than a three-paragraph post by bloggers like me but here goes.

First, it reminds me of the work "Some Thoughts About the Social Implications of Accessible Computing," by Multics developers E. E. David and R. M. Fano, circa 1965, that foresaw the future of cloud computing long before that term was invented to tout stock. The full connection is explained in this post I put up in March 2008 but the idea is the same as Ozzie's:

Build it and they will come... but before you build it, figure out what they want.

I get this feeling from his 1939 World's Fair references and other Carl-Sagan-type flights of prose. Actually to some extent, Ozzie's "future" memo looks back.

Second, and as an example of his looking back, I reject the idea that the technical press is trying to divine from Ozzie's writing that he is declaring the end of "desktop computing."  Ozzie doesn't say that. He says the era of a PC on your desktop is "winding down."  Of course, it is winding down.  Although "the dream" was basically about a PC on your office desktop, now there's one on almost every surface you own or pass by: your desk at work, your desk at home, your kitchen counter, your school laboratory's workbench (instead of a big honking Unix workstation), your retailers' checkout area (instead of a big honking cash register), and on and on.  The market saturation upper bound for what I'll rename fixed-surface computing is probably somewhere around the population of the world times n, where n is more than 2. If because of some collossal foul up in Microsoft development and marketing, Microsoft does not continue to own this still-growing market for fixed-surface computing and all its future migrations and permutations, it will still be "desktop computing."

That leads to the third point.  Ozzie is asking what follows "desktop computing."  From an IT investor's perspective, the question can be framed in three pieces:

• To what extent will a percentage of that upper-bound number of fixed-surface computers, whatever it is, be replaced by mobile-computing devices with natural interfaces (instead of a few hundred million of those weird little virtual keyboards that only a pianist can work that we have today)? Versus...
• To what extent will such mobile-computing devices also enable new applications that fixed-surface computers could not enable? In goods in transit (computers, not just tags)? In cars in motion in sight... legally? In your vest pocket? Inside your eyelids? Etc.? Etc.? And...
• To what extent could more traditional immobile fixed-surface PCs be used in new applications not tied to a person? At every intersection? On every highway milepost? Etc.? Etc.? (As you can see, my limited imagination is restrained by the fact that I just returned from a long trip by automobile.)

In other words, and I don't think I am disagreeing with Ozzie, you ain't seen nothing yet.

(Note-- and this is the reason Ozzie's memo needs more considered thought.--nothing I have written above or that I've seen in the press even begins to scratch the surface of server-side-computing's future, another whole continuing IT investment opportunity basically bounded by the sum total of all information -- right or wrong -- ever created in the history of the world.)

-- Dennis Byron

(no financial interest in companies mentioned)

Red Hat's Just Like IBM, Microsoft, Oracle, salesforce and SAP

James Whitehurst, Red Hat (RHAT) CEO, received a lot of IDG-based press during the week of October 18 about a trade-show keynote speech in which he reportedly said

""The business models between customer and vendors are fundamentally broken" when it comes to enterprise software." 

I did not attend the conference so I’m depending on these IDG press reports for the following opinion. (Truth in advertising: I am a former IDG employee so I trust the reports completely. This is an exception to the rule of not believing what you read in the newspaper, or at least the technical press.)

I think Whitehurst’s opinion is important more because of his former role as COO of Delta than for his current role at Red Hat. In particular, this quote rings true based on my demand-side research:

"Vendors have to guess at what [customers] want, and there is a mismatch of what customers want and what they get. Creating feature wars is not what the customer is looking for."

On the other hand, some of Whitehurst’s reported logic (possibly reported incorrectly) is shaky based on my supply-side research.  The press says his opinion about a broken business model is based on the fact that Red Hat is an “open-source software” company, and that others sell "commercial" software or have a different “software sales model.” 

I reject “open source” as an enterprise-software business model or sales model category. But if it is, isn’t Red Hat the only company in it?

Whitehurst’s logic flaw, if reported accurately by IDG, is that Red Hat’s business and sales model is the same as every other leading enterprise software company. 

• The very largest enterprise software companies -- IBM, Microsoft (MSFT), Oracle (ORCL) and SAP in alphabetical order -- have multiple-revenue-stream business models:
  • perpetual and various types of limited license revenues
  • subscription maintenance revenues (remember: software maintenance is about content not fixing things)
  • revenues from services that combine the license and maintenance with administrative support and the necessary hardware horsepower (so-called SaaS but increasingly also mislabeled as cloud computing)
  • revenue from services that combine SaaS with outright business-process services such as administering human resource benefits programs (so-called BPO)
• Large suppliers such as salesforce.com (CRM) license and sell only via the SaaS method while suppliers such as ADP use the BPO approach almost totally.
• Red Hat – last time I checked – makes its numbers via the first two methods (perpetual license with subscription maintenance), with the gotcha that its licenses are almost 100% made up of open source terms and conditions (Ts&Cs). But, oh by the way, IBM, Microsoft, Oracle and SAP also license some of their software with open source Ts&Cs.
• All of the above software suppliers -- and thousands of others -- also provide professional services such as training or implementation support for a fee.

If Red Hat and Mr. Whitehurst have something against these multiple revenue-stream or software-sales models – plural – it sure isn’t doing anything about it.

And there is an even bigger logic flaw in the press reports I saw (I only linked to one above but there were dozens). Whitehurst supposedly said

“"Why are costs of IT going up when the underlying costs to deliver those services halves (sic) every 18 months?" The cost of computing should come down, he reasoned, thanks to improving processing speeds and storage capacities. New, more powerful development tools and frameworks should also ease the cost of deployment. Yet IT expenditures continue to go up by about 3 percent to 5 percent a year.”

Unless Red Hat and the other leading enterprise software suppliers and his and their customers and clients are cutting their employees salaries in half every 18 months, processing speeds and storage capacities are not going to make much of a difference. Unless the taxes on the buildings they own or the rent they pay is being cut in half every 18 months, all the programming frameworks and “lower bug density” in computer-science nirvana doesn’t matter.  Along with software spend, hardware and storage costs add up to the $1.4 trillion IT spend mentioned elsewhere in the press reports. 50%-60% of that is annual subscription maintenance on hardware, software and storage and another 20%-30% is user add-on and functionality upgrades (so someone must like the feature lists).

But that entire $1.4 trillion number only accounts for about 10% of the cost of deploying IT in support of business processes. If you really want to cut costs, license and sell software that administers itself and runs on abaci.

– Dennis Byron

(no financial holdings in Red Hat but I have recently done research for Red Hat some of which is reflected in this white paper possibly still available on Red Hat's web site.)

Software Top 100 Ranking Shows Market's Dynamics, Lack of Maturity

Balder Verberne of softwaretop100.org, the research house based in The Netherlands, released his annual list of the top 100 software companies in the market worldwide in October. John Desmond of Software 500/Software magazine fame came out with his annual list recently too.  I like both lists -- and both researchers -- but I am partial to the purity of the Software Top 100. The Software 500 has added non-software-specific services revenue to its ranking methodology in the last few years.

As I've noted elsewhere, as a guy who used to be able to build such lists in his sleep when working for IDC, I like to give such research a quick once-over to see if I agree with the rankings and relative sizes attributed to each enterprise software provider on the list.  Everyone -- including the big quant houses such as Dataquest and IDC -- agrees that IBM, Microsoft (MSFT), Oracle (ORCL) and SAP are the leaders in the software market. That's in alphabetical order, not actual ranking. 

But I'd have given Oracle a little more revenue than Balder gave it in both 2009 and 2010 and I'd have given SAP a little more credit. I think SAP grew a little in 2009 as opposed to Balder's estimate that it deflated a little. I suspect our differences vis a vis SAP are exchange-rate methodology differences, always a problem in estimating SAP. Relative to Oracle, I suspect the softwaretop100.org group may not be picking up the small but growing amount of Oracle software revenue in its OnDemand segment.  In fact, I think Oracle is probably already second to salesforce.com (CRM) in terms of software as a service (SaaS) revenue.  (Or possibly Balder does not "backcast" in revenue from the acquired company in the IDC fashion, but uses the suppliers' GAAP reported numbers. That's OK for rankings but causes all kinds of problems with forecasting market size using the totals resulting from the rankings.)  

Further down any such list is where it gets interesting.  softwaretop100.org has added Ericsson (ERIC), Nokia (NOK) and the like this year and ranks them in the top 10.  Although they deserve to be somewhere on the list (one of my best business trips of all time was the four days I spent in Helsinki at Nokia HQ), I'm not sure I'd credit it with software being 25% of its revenue. Software market lists should really measure software sold/bought as software to truly measure market dynamics.  But I admit that I have not spent a lot of time studying the issue vis a vis the telecomms.

Primarily for me looking at such lists as the Software Top 100 and Software 500 these days is an "in memoriam" exercise to see which companies are no longer with us as compared to the year before.  In the 2010 softwaretop100.org list (rankings for 2009 of course), the only company we say goodbye to is SPSS by my quick reckoning.  That may be a sign of advanced software market maturity but in fact we already know that in 2011 we will say goodbye at least to Sterling Commerce, Sun, and Sybase.  Maybe software companies are now being acquired in reverse alphabetical order.

By the way, in my opinion softwaretop100.org did a good job of picking up the Sun 2009 total. After the Oracle acquisition was announced there was no more data from Sun but a pretty low software -- and overall total -- has now been confirmed in the proforma section of the Oracle 10-K. 

-- Dennis Byron

(no financial interest in companies mentioned)

Ozzie Out; The End of "Software And Services" at Microsoft

In the long history of "you're out on your ass" press releases from major companies, Microsoft's (MSFT) announcement of the departure of Chief Software Architect Ray Ozzie is the strangest one I've ever seen.  

But what it means for Microsoft investors is bad news. 

As described on this blog post I wrote for Research 2.0 in February 2007 (and which appears on SeekingAlpha as an anonymous submission and also appears somewhere in the archives here on my own site)

"Basically, Microsoft has to decide whether it wants to be a utility, or just be the guy that makes the dynamos. Or the electric motors. Or the circuit breakers. It is important to remember that Ozzie's plans are still not necessarily Microsoft's plans. There is also a Ballmer plan and Bill Gates hasn't 'retired' yet. (Author's note: remember this was written in 2007)

"I think in deference to the others Ozzie uses the term "software and service" rather than software as a service, even though that (author's note: "that" refers to the words "software AND service") would be a step back to the 1990s. If Microsoft tries to straddle that fence (be both a service provider and a technology supplier), its investors will be the ones that feel the pain in the middle."

I've repeated all the illogic and investment problems behind the 'and" in the term software and services in multiple places since, usually right after the annual Microsoft financial analyst day each July. I wrote a note to myself to ask why Ozzie didn't speak this year but then went off on vacation myself and didn't follow up.  Now we know the answer.

Microsoft never made the leap to software as services that it needed to make and Ozzie was trying to architect (which is what chief architects do after all). I fear Ozzie's departure to pursue other interests or retire or play video games or whatever Steve Ballmer is talking about in this weird announcement means Microsoft chose wrong.

This is the high-level departure that Larry Ellison should be commenting on.  Please, Larry, send me an email about it rather than wasting one on the New York Times or Wall St. Journal.

-- Dennis Byron

(no financial interest in companies mentioned)

Is Fable the Best RPG Ever? No, RPG/400 Was!

Ok, it was pretty stupid of me on October 3 to get hooked by the headline "Was Fable the Best RPG Ever?" in PCWorld.

No, I said.  First of all, there is the Clintonian answer:

"It depends on what the meaning of best is."

But, second, clearly the best RPG was IBM III and its sibling RPG/400.  In fact, what the hell was Fable RPG? There was a DEC RPG and a Honeywell RPG and Data General RPG. There probably was a Wang RPG. But I don't remember a Fable RPG.

Anyways IBM RPG circa 1965, especially III and 400 circa 1985, spawned what is today the $30 billion a year ERP market (just in enterprise software sales, not counting systems and services drag of 3X to 10X depending on the year).  Actually the term ERP didn't exist then. It was MRP for material requirements planning and then MRP II for manufacturing resoruce planning.

The term ERP was coined by Gartner around 1995. When I joined IDC in 1997 from Datapro where I was the "the ERP analyst," I was told not to use the term ERP because Gartner coined it. No kidding.

The ERP market started with the granddaddy of them all, RPG-based IBM MAPICS, still being marketed and supported I believe by Infor. Marcam was designed as ERP for discrete manufacturers so a Mapics distributor named MARCAM wrote a version for process manufacturers also in RPG.

And there was J.D. Edwards World and predecessors, which now can be found somewhere in Oracle (ORCL), SSA BPCS before BPCS VI (also in Infor as is Marcam but not necessarily the RPG version), initial Intentia products -- now part of Lawson (LAWS) -- and many other early leaders of the ERP market circa 1985-1995 were also originally written in RPG.  Actually a better term would be written dependent upon RPG.

This 2007 article says there are still 3000 Edwards World users. That's based on some Internet browsing -- so don't take it as gospel -- which also says MAPICS added a Java version to go with its RPG version around 2000 whereas Intentia ported all of its products to Java and left RPG behind shortly thereafter.

Both SSA and Marcam infamously tried to write object-oriented (OO)-based versions of their RPG products in the late 1990s, with Marcam just falling short as often happens because OO was so new at the time and SSA falling short spectacularly with SEC investigations and founders supposedly hiding in Hong Kong or some such place.

Given its role in ERP -- not to mention all the in-house developed RPG software for the AS/400 and predecessors -- the IBM Report Program Generator III has to be the best RPG ever.

So I whipped off my answer to PCWorld. 

Boy, do I feel old... and stupid.

(Sidenote: Before I get a deluge of email from people even older than me, I consider COPICS to be the great granddady of ERP.)

-- Dennis Byron

(no interest in companies mentioned) 

Dell/Perot Caught in Massachusetts 2010 Election Crosshairs

On October 1 I wrote about the leg up that Dell/Perot had as Obamacare forces the roll-out of over 40 IT/help-desk-heavy healthcare insurance exchanges across the United States. This is an opportunity for HP (HPQ)/EDS and IBM as well but Dell/Perot gets the advantage primarily it built and "mans" the exchange here in Massachusetts, called the Commonwealth Connector, that is part of Mitt Romney's signature healthcare insurance reform effort. The 2006 Romneycare legislation is widely considered to be the model for Obamacare (more by Obama than by Romney), which is scheduled to be implemented throughout the United States between now and 2019, primarily in 2014.  

Current political crosscurrents here in Massachusetts however are illustrating one of the downsides of that exposure (which, as I also wrote on that October 1 post, is unlike the situation in Utah because Utah is not outsourcing the IT and help-desk work to another state to the extent Massachusetts did). So today we have Texans or people from wherever (if they are Indian as I hypothesized in my earlier post, their accents are great) that "man" the help-desk/phone-bank boiler room trying to defend the fact that individuals, mostly small business people, in Massachusetts cannot buy healthcare insurance at all for the next three months. 

And it's apparently not just a matter of not being able to buy it from the Connector.  As this Massachusetts Department of Insurance (DOI) web page explains, the problem applies to all insurers and brokers. In Massachusetts that's mostly non-profits but UnitedHealthcare (UNH), Centene (CNC) and other insurers in whom investors would be interested also sell to individuals.

A similar situation occurred in April 2010 when Democratic Governor Deval Patrick turned down rate increase requests by the insurers and froze the market for individuals and small business people trying to get healthcare insurance for two or three months.  Patrick is a running for re-election and his effort was demonstrably political grandstanding as subsequent DOI administrative court rulings -- in other words, from his own employees -- said that he both did not have the right to turn down the rate increases and that the increases were actuarially justified.

This time it's the Republicans that are doing the political grandstanding, particularly a Republican named Karyn Polito who is running for State Treasurer and is apparently trying to misdirect news about alleged ethics violations in order to save a foundering campaign. (Believe it or not, and despite what you read about us here in the Bay State, ethics violations might harm her here in Massachusetts.)  The story is that the Massachusetts Democrats who passed the law prohibiting individuals and small business people from buying healthcare insurance for three months really didn't mean to.  They seem to have adopted the U.S. Congress mentality of passing a law first and then reading it later. They are trying to correct the problem in an omnibus technical-corrections bill but Polito has blocked passage for a week so far in order to get publicity for standing up to the Democrats.

Both situations -- both the one in April and the current brouhaha -- illustrate how unfriendly over-regulated, over-mandated, over-governed Massachusetts is for small businesses.  Fortunately you can't beat Cape Cod in the summer or the Berkshires in the autumn.

-- Dennis Byron

(no financial interest in companies mentioned)

Did Larry Really Mean to S*#t All over Leo?

Taking my cue from Mike Fauscette (who basically said in his recent post that you can't consider yourself an enteprise-software/IT blogger and not comment on this story), here's my two cents on Leo Apotheker being appointed CEO at HP (HPQ). And more important, on Larry Ellison's opinion of Leo Apotheker being appointed CEO at HP.

I believe: 

"It's the HP board that Larry is s*&#ting on, not Leo." 

(Or do you spell 's*&#ting' with two t's?)

So off I went to hp.com to see who was on the board.  The first thing I found strange was that HP has a page, right beside its "Board of Directors" page, called "Former CEO's." 

Really! 

Yep, there's Carly. And there's Mark. Really! Maybe Larry's right!

I am guessing this HP web page started with the idea of some PR guy to give tribute to Hewlett, Packard and Lew Platt.  But guys, you could have titled it "In Memoriam" and left off Carly and Mark. I don't think John Young would mind.

But on to the subject at hand, to a Valleyboy like Larry the problem at HP is clearly that HP's board does not include a lot of Valleyboys. Of course, there is Marc Andreessen, Netscape founder and inventor of Mosaic but for the most part the board is made up of real businesspeople as opposed to those trapped in the 101 bubble. Maybe there's some bad blood between Andreessen and Larry.  Andreessen's also on eBay's and Facebook's boards and was into the "cloud" more than 10 years ago, back when Larry was still saying that the client/server software architecture would never take off.

Other than Andreessen, there's John Joyce formerly of IBM Global Services via PWC, a guy from Verizon (VZ), a woman from Nokia (NOK) and another who was a TV network executive, a guy who spent his career at a chemicals/specialty-materials manufacturer that is now a subsidiary of Dow, the chairman/CEO of McKesson (MCK), a guy who is also a director of General Mills (GIS), the former chairman of Wachovia (ouch!), and the guy who appears to have invented the legal maintenance organization (LMO), a sort of HMO of lawyers.

In other words, it looks like the HP board is made up of HP customers instead of Silicon Valley geeks and the financial types that suck up to Silicon Valley geeks hoping for insider information. And it looks like this group of proven business managers brought in a proven business manager.

So in the end, as is true of most Silicon Valley navel gazing, this is much ado about nothing. But it still could be a problem for Apotheker; time will tell. In Palo Alto, no one wants to work for a company that doesn't play the game. And the valleywags are already at work making this appear to be the worst corporate idea since bringing John Scully into Apple (APPL).

That's apparently where former Oracle (ORCL) President Ray Lane as HP Chairman comes in. (Unlike HP's "Former CEO" page, there's not enough room on Oracle's web site for a "Former Presidents" page.) As Managing Partner at Kleiner Perkins Caufield & Byers, he's the prototypical Valleyboy although prior to 1992 he also worked at boring places like BoozAllen, IBM, and closing the loop back to HP... EDS. 

Lane's job was to make sure people like Larry didn't s#*t all over Leo. Oh well, "C'est la vie," as Apotheker would probably say. 

Solution: Pack up "the garage" and move the whole HP shebang to Salt Lake City or Albuquerque. What would it hurt? According to Barbara Boxer, Carly already laid off all the HP employees in California.

-- Dennis Byron

(no financial interest in companies mentioned)

Obamacare: And the Winner is... Dell/Perot/Texas/(India?)

The Wall St. Journal (WSJ) reports September 30, 2010 about the information technology (IT) and other shortages that U.S. states will face implementing the Obamacare health insurance industry mandates between now and 2019. That's especially an issue in terms of the many IT and call-center aspects of the Obamacare kickoff in 2014. 

The WSJ news should be good for DELL and its services arm Perot Systems as well as for employment in Texas and any other geograpy to which Perot outsources IT and call-center work. (That's ironic given the Texas state government's and Texas citizens' opposition to Obamacare.)

Dell/Perot has a leg up especially in terms of the need explained in the WSJ to build 40-plus insurance-brokerage exchanges from scratch in the next few years because Perot had already done that here in Massachusetts, the U.S. state often termed the test lab for Obamacare. Utah, the other state noted for its pioneering health insurance exchange work, appears to have done its systems integration inhouse based on services and enterprise software from eHealth (EHTH), Chicago-based bswift and Utah-based Health Equity. Apparently there was less opposition here in Massachusetts to outsourcing employment to other states. This sort of thinking might change based on upcoming U.S. elections but Massachusetts' approach to healthcare insurance reform is more popular among the U.S. federales right now than Utah's.

I would expect IBM and HP (HPQ)/EDS (that other little Perot start-up) to be fighting tooth and nail for this business state house to state house as well.

Microsoft (MSFT) might be a big winner too although I haven't dug that far back into the supply chain to say for sure.  As recently as three or four years ago, Dell was such a large Microsoft customer (more than 10% of Microsoft revenue) that it was called out in Microsoft's SEC filings.  And despite open source propaganda to the contrary, Massachusetts state government is a major Microsoft shop.

One of the Utah-state-owned-insurance-broker software providers, bswift, uses Rackspace (RAX), providing another Texas connection.

-- Dennis Byron . 

(no financial interest in companies mentioned)