Description

Despite Dell’s long track record of consistent growth, cost leadership, and high returns on capital, the stock trades at its lowest valuation of the last 9 years. Dell has stumbled over the past 18 months due primarily to inferior products, declining service, better competition, and saturation in the U.S. Better competition and a mature U.S. market are here to stay, and that (combined with needed investments in service) likely means Dell’s profitability will remain near the low end of its historical range. But Dell’s products are improving and will be as good as the competition (if not better) within a few months in every PC market segment. Dell is finally taking steps to improve service. And there is still room for growth in servers, international markets, niche PC market segments such as gaming, and nonPC product lines such as printer cartridges and TVs.

Furthermore, as the PC industry continues to consolidate, the “toll” that many software makers must pay to have their products bundled with new systems will continue to increase and this will provide growing revenue streams to the largest PC vendors Dell and HP, widening their advantage over smaller PC makers (the Google deal is just the latest example and has helped escalate the toll cost). Though Dell’s cost advantage has shrunk in recent years, the build-to-order PC model will always have at least some cost advantage over the indirect model of the competition, which should allow Dell to recover from those occasions, such as the last 18 months, when it makes a series of product or service mistakes.

Founder and largest shareholder Michael Dell seems to believe Dell’s performance (and stock price) will improve as he just purchased an additional $70 million in stock at the end of May. This is just the third time Michael Dell has purchased stock in the open market – and the first two were well timed (early July 1992, late September, 2001).

Dell was posted to VIC just 6 months ago, and it appears that quite a few value investors have already bought Dell stock – so why post this idea again? Other analysis of Dell I have seen avoids product detail, in spite of the fact that the information is widely available and has much to do with Dell’s lagging results of late. Product analysis led me to believe that Q106 and Q206 sales and/or earnings would be far below what most analysts were expecting, which is why I waited to purchase Dell stock until after Q1 results were reported. I also believe the risk/reward for this stock has improved since the last write-up due to new service initiatives, a product line within one quarter of dramatic improvement, and of course, the lower stock price – so low that Michael Dell bought a bunch.

While calendar 2006 will not be a good year ($1.00-$1.50 EPS depending on how aggressively Dell cuts prices), I believe that both sales and earnings will rebound in 2007 ($1.50 to $2.25 EPS) and will then grow by at least 10%/year over the following 5 years.

PRODUCT LINE

Though Dell has lagged the competition in price/value for most products over the past year, they will again be in the lead or at least competitive by Q406. The biggest problem in Dell’s lineup is that AMD chips have provided much better performance and power usage characteristics over similarly priced Intel chips (desktop and server) during the past year, yet Dell does not sell AMD-based systems.

In order to compensate for the lack of AMD (and for other product line issues as well as the improved cost structure of competitors), Dell has flip flopped between maintaining profit margins at the expense of sales or increasing sales at the expense of margins. April 2006 was a month in which Dell went to price cutting extremes. This had the obvious impact on margins which led to the negative “surprise” Q1 announcement in May. From my tracking of Dell’s U.S. web site, I see that Dell continued to price aggressively in May and June in the U.S. consumer and small business segments. Dell’s Q2 is thus likely to be poor (in either sales growth or earnings growth depending on the pricing strategy chosen), particularly in those categories where savvy customers will reduce or completely halt purchases until Dell ships systems based on the new Intel chips.

Intel’s new chip families being released Q3 (and a variety of other minor factors) should allow Dell to resume profitable growth –not as good as past growth rates – but better than what the market is expecting. Most of the new chips have significant advantages over similarly priced AMD chips. The new chips will be supply constrained at first, with Dell having preferential access given that they are Intel’s largest customer. Dell also benefits by receiving steep price cuts on current Intel chips many weeks before others get the same pricing due to its smaller inventory. In other words, for the first time in nearly 2 years, Dell should benefit from its Intel focus and a lean inventory business model which thrives on rapidly dropping component prices.

Following is a detailed analysis by product category of why Dell is behind but will catch up or seize the lead within a few months.

SERVERS

Server market growth has slowed over the past year but will likely be a fast growing segment over the next decade as the shift towards web-based storage and applications continues. Thanks to a shift away from expensive systems to sub $25,000 systems based on the x86 architecture, Dell steadily gained market share through the end of 2005 (competitors sell a mix of higher and lower priced server systems but Dell only sells the low priced x86 based systems). However, within the x86 segment, AMD’s chips have been vastly superior to Intel’s server chips over the past year, and this caused Dell to lose server market share in Q106 for the first time and miss out on an opportunity to gain market share more rapidly in 2005.

Increasingly, the biggest bottleneck in server clusters is wattage. Energy-hogging servers overtax the local electrical grid and often require extensive cooling to run reliability – and of course this results in huge electricity costs, sometimes with annual costs approaching the initial server cost. Given that the inferior energy efficiency of Intel server chips compared to Opterons from AMD, Intel-based Dell servers are inferior solutions for large server clusters and are increasingly shunned in favor of Opteron-based solutions.

All this will change Q306 as Intel just released its next generation server chip family (Woodcrest) which boosts performance while dramatically lowering power requirements. Early testing indicates these new chips will have an advantage over Opterons in both performance and power usage (and a large advantage for dedicated web servers), and Intel has indicated that they will be pricing aggressively to regain lost share. Furthermore, Dell will be launching a server product using AMD’s Opteron in a product niche that Intel’s server chips will continue to lag (4-way servers).

Dell will have preferential access to these chips which may be in short supply as they are being rolled out. I therefore expect rapid growth in server market share for Dell as machines based on these new chips begin to ship in volume. It might take many months of customer testing and ramped production by Intel before sales begin to really take off.

GAME MACHINES

Though a relatively small portion of the overall market for PCs, gaming systems command higher prices and margins than the commodity machines used by office workers, and this segment is growing fast. They also tend to generate much more intense brand feelings (negative or positive). Dell has been derided by gaming enthusiasts in recent years because their XPS machines are not stylish, are loaded with bloatware (preinstalled software that causes stability and performance issues), don’t use AMD chips, and some XPS models are not good for playing even rudimentary games.

In 2006, Dell addressed (or will soon address) all of these issues. Dell now has a stylish, AMD offering and popular gaming brand via the Alienware acquisition (which adds $200 million annual revenue to the $1.3 billion XPS revenue). Dell has been introducing more stylish and gamer oriented machines in the XPS line. In April they began to offer limited bloatware as an XPS option, and they have overhauled the pre-installed software on the other XPS models to remove most of the stability and performance issues and much of the desktop clutter.

Most important, the new Intel chip families being released later in the year will have better price/performance than AMD’s offerings. Consequently, I expect Dell’s XPS systems to garner more respect among gamers and rapidly increase market share.

If you want to see for yourself just how bad Dell’s XPS line was half a year ago and how committed Dell is to improving, I strongly recommend reading through the December 16, 2005 Dell XPS 400 review, the May 24, 2006 “Dell Strikes Back” article, and the June 12, 2006 XPS 200 review on the following excellent review website (also read the 2 eMachines reviews to see the kind of competition Dell is up against for low end desktops):

http://consumer.hardocp.com/


DESKTOPS

For most of the past year, at nearly every price point, the price/performance ratio for AMD desktop chips has been superior to that of Intel chips. Consumers, without necessarily understanding much about what chips are inside, have been shifting their business to the best value systems and this has resulted in sales of Intel-based machines growing more slowly than those based on AMD. For example, in the U.S. consumer retail channel, AMD systems outsold Intel systems by over a 3 to 1 margin in the first half of 2006, the highest ever. The impact of Intel-only on enterprise sales has been less severe, as many IT managers care more about standardizing to reduce maintenance costs than processor performance.

As Intel’s largest customer, Dell receives rebates and other advantages from Intel, so they are the low-cost leader for systems with Intel inside. Nevertheless, competition with AMD-based systems has been intense. Dell has usually chosen over the past year to harvest higher profit margins on desktops, resulting in Dell systems priced above AMD-based systems of equivalent power at most price points, and resulting in stagnant market share. The superior price/power/performance characteristics of the new Intel Conroe chip will likely drive a return to Dell steadily taking desktop market share, as usual.

Even before Q3, Dell should see at least a little improvement. In April, Dell began shipping most mid to high-end systems with dual core chips from Intel (or at least offering an inexpensive dual core option). Dual core systems allow users to continue working without interruption while doing heavy multitasking (think downloading, backups, virus checks, or recording a TV program while using a computer to do normal work).

Until April of 2006, dual core systems have been a nice technology but too pricey for most buyers. But Dell is bringing these dual core systems into the mainstream with very attractive price points – some mid-range U.S. systems (with Pentium Dual Core 820, 1GB RAM and 19” monitor) below $600. These dual core systems are the one desktop offering from Dell in recent memory that clearly offers more bang for the buck than competitor offerings (AMD dual core chips are good but expensive – they make more sense for higher end rather than mid range systems). While probably good for market share, these systems will put pressure on profit margins as the pricing on some of these systems is too low to be purely accounted for by special chip pricing from Intel.

These Intel dual core systems are energy hogs. It may not matter to PC buyers who are not over-clocking gamers, server cluster managers, or environmentalists – but on the other hand it might matter to those buying Media Center machines that are noisier from the high fan speed required to cool these heat-spewing units. This issue will be resolved in August when Intel’s new energy efficient Conroe chip family (also dual core) begins to ship in quantity. Intel will be aggressively pricing the new dual core chips to regain share lost to AMD over the past couple years and Dell will be the biggest beneficiary given its relationship with Intel.

In considering systems for purchase, current mid-range Dell desktops (5150 for small business, e510 for consumer) are reasonable to recommend as they are expandable and are the least expensive dual core systems available. But Dell’s low-end systems lag the competition. A lack of PCI x16 slot, a slow processor, and a dated supporting chip set, make the Dimension E310/3100 product line a poor choice for most buyers. The lack of the PCI x16 makes this system incapable of being upgraded to run the full Windows Vista (Aero interface) experience and totally unusable for modern games. Many competitors offer AMD-based machines at lower prices with the PCI x16 slot and faster processors. For example, the eMachines T3418 system with 17” CRT is sometimes advertised (including rebates) for $350, and the more powerful T6528 for $400 (consumer.hardocp.com reviews both units). Both are capable of running modern day games at very low speeds, and the T6528, if combined with more RAM and a graphics card upgrade for $200 can be a respectable game machine. I expect Dell to replace the Dimension E310/3100 product line with a new line that has the PCI x16 slot and faster components so that consumers who want to spend less than $500 on a system capable of gaming or fully running Windows Vista won’t be forced to shop elsewhere.

Yet another problem with Dell systems is the preinstalled software, commonly referred to as “bloatware” because it causes systems to be slower and less stable. IT managers have to go through a 1-2 hour cleanup process on many models of Dell systems to get rid of all the bloatware, before they even begin to configure the system. For consumers who don’t know better, the poor stability of many Dells out of the box can be quite frustrating. And the number of tech help calls generated by bloatware issues must result in significant service costs and unnecessary complexity. While other companies have bloatware as well, for some reason Dell has recently done a poorer job of testing preinstalled software for stability and thus shipped systems which are less stable and slower than most of the competition, even though the hardware is usually as good or better. For this reason alone, one might recommend HP or eMachine systems for most consumer purchases. This issue is obviously fixable and as I noted in the gaming section, Dell has begun to fix it.

Perhaps the most serious issue for Dell in the long-term is that sub $500 PCs have become the norm for consumers. Competitors are willing to accept very low margins on these machines, and Dell is unable to compete effectively without also sacrificing margins. Higher fuel costs can’t be helping as shipping costs are a sizeable fraction of low-end system cost. The launch of Windows Vista may help temporarily next year as requirements to fully run Vista may prompt many consumers to buy more powerful machines. But long-term it is not clear how Dell will compete with the sub $500 consumer PC – and this is partly why I believe Dell’s net profits will settle at a permanently lower level of around 6%.

It is worth noting that many U.S. businesses are at a lull in their upgrade cycles (many did upgrades just before Y2k, then another round in 2004 to 2005) – and this impacts Dell more than other PC makers because Dell has much greater market share of U.S. business PC sales than its competitors. I think it is also possible that some IT managers have cut back on systems purchases as they await systems based upon the improved Intel chip families to be released in Q306.

NOTEBOOKS

Notebooks are the only area where Intel’s chips have remained superior to AMD over the past year. AMD has used price to try to compensate, as AMD wants to provide competitive chips in every category. Not surprisingly, Dell has done better with market share lately in notebooks than desktops. But according to management, Dell has less of a cost advantage in notebooks because assembly is mostly by third parties (the same ones used by competitors). So the more overall demand shifts towards notebooks, the more difficult it will be for Dell to retain its historical margin and market share growth advantages over the competition.

In May, Dell introduced a wide range of new notebooks that have dual core chips, at very aggressive price points. Some of these designs are more attractive than most notebooks Dell has shipped in the past. I would guess that Dell will begin to see reaccelerated market share gains with these new notebooks, but will likely experience margin pressure in the short term from its aggressive pricing.

The new Intel chips to replace the current line will be an improvement in price/power/performance but much less so than for desktops and servers. And Dell will continue to have a smaller cost advantage in notebooks than desktops versus the competition. So of all the different segments, I expect notebooks to show the least improvement over the next year.

AMD VERSUS INTEL

As described above, Dell has been handicapped during the past year by not having AMD chips, especially in Servers and high-end gaming desktops. Below I reiterate why the Intel versus AMD factor is about to change into an advantage for Dell.

Dell recently acquired Alienware, and this plugged a hole in their desktop gaming lineup. After a year of slow price changes, Intel is rapidly dropping chip prices (with more to come in July), and this makes many more of Intel’s chips competitive or superior to similarly priced AMD chips. Dell gets the benefit of dropping prices weeks before most of its competitors due to its smaller inventory (days as opposed to weeks).

Most importantly, Intel is launching new chip families which, by the end of Q306, should cause Intel-based systems to exceed the performance of AMD systems (speed and heat dissipation for gaming, good ratio of power usage to performance for servers, and generic price/value relationship for the rest). This performance advantage may come in particularly handy as Windows Vista is launched and PC Buyers want to insure their systems are speedy enough for Vista. Dell’s preferred partner status with Intel will thus be much more of a competitive advantage for Dell than is presently the case, as Dell gets preferential access and pricing to the new Intel chips, which will only be 25% of Intel’s production by year end (However, see the first item in the Risk section on the Dell/Intel partnership).

Generally speaking, healthy competition between AMD and Intel is a good long-term development for Dell as it should increase negotiating leverage over Intel in the long-term, regardless of who the current leader is in the AMD versus Intel horserace.

SERVICE

A variety of Dell service failures have been broadcast over the internet over the past year, damaging the brand. Even excluding the most spectacular service failure examples, many customers are frustrated with language-challenged tech support from India.

Dell is not alone with service problems. It’s just that Dell used to be the service leader. Over the past several years, Dell lost its large lead on the consumer side and its small lead on the business side, and refused to even acknowledge there was a problem.

Dell has finally acknowledged the problem (they are now talking about it frequently) and is taking several steps to improve service. Dell is hiring more tech support people, increasing the percentage whose native language is English, beefing up training, and has recently launched a service that allows tech support to “take over” one’s computer and fix it remotely. Dell is also beginning to reduce software bloat, which should lead to fewer tech support issues in general.

Dell has so far committed $100 million to improving service and is beginning to see signs of improvement – reduced hold times and improved experiences reported by some reviewers. It is difficult to predict, though, how long it will take and how much it will cost. There is certainly much room for improvement. I have personally experienced poor Dell tech support, including an experience earlier this month when it took three departments and nearly an hour to order a set of replacement disks for Windows XP and Dell drivers – and it was accidentally sent twice.

OTHER

Dell’s printer line is still in the investment phase. Printer products will contribute substantial profits as cartridge sales catch up to printer sales (though there is risk this may not work out due to growing cartridge competition from remanufacturers and refillers).

Some of Dell’s foreign market growth efforts are still in the investment phase, and will contribute much more to EPS growth when Dell attains critical mass market share and lightens up on investment.

As Dell market share grows and Microsoft and Intel’s monopolies weaken, negotiating leverage should tilt more in Dell’s favor over the next few years. Only one competitor (HP) is large enough to have similar negotiating leverage.

Dell is trying (and succeeding) to increase its market share in the service business, which they have historically left to others. Dell’s service market share is still low so there is still much growth potential.

THE NUMBERS

2006 will be a bad year for Dell and it is likely that EPS will range between $1.00 and $1.50 depending on Dell’s price aggressiveness. With Dell’s much improved product positioning towards the end of 2006, higher payments from preloaded software deals, and increased printer profitability (offset at least partly by better competition), I expect calendar 2007 to benefit with increased sales growth and a return to net margins at least as good as the low end of the prior 10 year range. This means net profits of at least 6.0% on $65 billion in sales. These numbers are not meant as a precise prediction but rather normalized sales and earnings upon which I expect at least 10% growth per share going forward (see prior VIC Dell write-up for a reasonable explanation of the 10% future growth rate assumption).

Net profit percent will be so low because the competition really has improved, industry growth is slowing, and Dell will probably need to continue to sacrifice some margin if it wants to keep gaining market share at rates similar to that of the 2000-2004 period. Dell has frequently indicated that their business model works best when they are gaining market share. Therefore, I believe Dell will emphasize a return to historic market share growth rather than a return to historic levels of profitability – and I think Dell is being forced to make this choice. Dell is also investing heavily in a variety of service initiatives in a bid to regain industry leading service performance, including lengthening the standard warrantee from 90 days to 1 year on many systems. I believe this is the right thing to do – but it will permanently reduce profits as a percent of sales.

If Dell’s stock price averages $30 over the next year and Dell continues to buy back stock at the present rate of about $6 billion per year, then 200 million shares per year will get retired, offset by about 50 million/year in new options issuance, for a net of 150 million/year. Based on these assumptions, average weighted basic shares in 2007 will be approximately 2.1 billion, and diluted shares around 2.2 billion.

Thus, normalized diluted EPS of $1.77 in 2007, leads to a valuation of $30/share if EV/E is 15 (assumes an average $8 billion in cash and short-term investments net of debt during 2007, or $3.60/share of cash), or $39/share if EV/E is 20 in mid 2007. The actual EPS could be anywhere between $1.50 and $2.25 in calendar 2007, but Dell will be looking a lot healthier as growth rates and profitability are improving at that time. For reference, today (with the stock price at 24.6), EV/E is about 16 based on depressed EPS of $1.25 for calendar 2006

Note that stocks of technology companies tend to respond strongly to shifts in market share gains, and Dell stock has responded similarly (most recently in a negative way). Assuming my product analysis is correct, Dell will be gaining share again by the end of 2006 and more strongly in 2007, and this will be a strong catalyst for the shares to become properly valued, if not overvalued.

SEE PART 2

Catalyst

Improved products (mostly from new Intel chips) and service will lead to resumption of steady market share gains.

More preloaded software deals and maturing international and printer business lines will increase Dell’s profitability and/or rate of market share gains, depending on the pricing strategy chosen.