Author Archive: Gary Kinman

October 12, 2007

The True Value of a Hosted Server

Now that I've ranted on a few accounting shortfalls for the hosting industry I'm going to rant once more. I think that the way hosting companies must book the value of their assets per accounting rules shortchanges hosting companies. Some basic rules of finance clearly show the likelihood that significant value is missing on the financial statements.

Let's consider a mythical server that costs the company $10,000 to buy and the company depreciates it evenly over 3 years. After year one, the value on the financials is $6,667. After year two, its book value is $3,333 and finally $0 after three years. Suppose that the company deploys the server for five years. In reality, after three years, the server's true value is certainly above $0, and the hosting company is shortchanged by not being able to reflect this value on its financial statements. Multiply this effect by thousands of deployed servers and you can see that there is significant value in hosting companies that just isn't found on the financial statements.

So how should we reflect the value of a server? I would propose the use of a "capitalization rate" or "cap rate". This is a common method of appraising real estate and the formula is simple: take the projected net cash flow over the next 12 months and divide by the cap rate, and that's the value. So, what would happen if we applied this to a server?

Looking at our mythical $10,000 server above, for simplicity's sake, let's ignore any allocations of the switches, routers, generators, HVAC, etc., needed to operate it. Let's also assume it produces net cash flow of $100 per month and will do so for 60 months. Its 12 month projected net cash flow is $1,200. We would divide this by the cap rate to find its value.

Naturally, the next question is "what do we use for the cap rate?" For a given investment, the cap rate is the lowest return that an investor will accept for the given risk of that investment. In our server's case, the $10,000 investment produces a return of $1,200 per year. How much would an investor need to invest in lower risk alternatives to get the same return? For a risk-free investment of the same 5 year duration such as a 5 year Treasury Note at 4.25%, you would have to invest $28,235.29 to get $1,200 per year in return. If we use 4.25% as the cap rate in our scenario, the value of the server becomes $28,235.29. However, investors in hosting companies generally look for returns far above 4.25% and these returns are not without risk, so this is not the appropriate cap rate. For simplicity's sake, let's assume that the hosting company investor's minimum acceptable rate for the investment is 10%. In other words, if his investment in the hosting company was expected to return less than 10%, the investor has other lower risk options to invest and get a 10% return and he would not invest in the hosting company.

So if we use 10% for the cap rate in our mythical server scenario, the true value of the server is $12,000 ($1,200 / 10% = $12,000). As long as the 12 month projected net cash flow stays above $1,200 then that value holds constant. Check out the graph below to compare the value of this server from both the cap rate perspective and the accounting rules perspective over the five year life.

From month 36 to month 49, there’s a $12,000 difference in value between the two methods. If a hosting company has a thousand servers like this, that’s $12 million in value that isn’t reflected in the company’s financial statements. That’s huge.


September 18, 2007

Current Ratio Punishes Hosting

For the third (and final!) installment in this likely sleep-inducing trilogy of hosting and accounting blog posts, we'll cover Current Ratio and how it doesn't treat hosting companies fairly. Bear with me – this rant may run a bit longer than normal.

Current Ratio is easy to compute – simply go to the balance sheet and divide current assets by current liabilities and voila! You have the Current Ratio. OK, so what does it mean?

So why is this unfair to hosting companies? Well, where does most of the cash of a hosting company go? Into servers and networking gear! But guess what? These don't fall under current assets on the balance sheet and thus are excluded from the Current Ratio calculation. As a result, I'd wager that most if not all hosting companies have at some point been in the position of current liabilities being greater than current assets, where conventional wisdom says "the company may have problems meeting its short term obligations."

How does this hurt hosting companies? Suppose the company could use some short term financing for a network upgrade. If they go talk to a banker about this the banker might throw up his hands and say "I can't help you…you're in financial distress according to your Current Ratio."

I would argue with the banker that this is not necessarily so. Traditional GAAP places servers and networking gear in the bucket of long term assets along with things like buildings, bulldozers, cranes, heavy machinery, etc. For a hosting company, this placement just doesn't make sense.

Long-term assets, or capital assets, are things that typically can't be reconfigured, can't be easily converted into cash, and are used for a long period of time. A hosting company's, buildings, generators, HVAC gear, etc., is rightly classified in long term assets. But servers and networking gear are quite different in that they exhibit more traits of current assets than long term assets. Check out this definition. It would take far less than a year for a hosting company to convert its server fleet and networking gear into cash and these assets are the key source of funds for day-to-day operations.

A manufacturing company gets to count its inventory in current assets, whether it is raw materials, work in progress, or finished goods ready for sale. A hosting company uses its servers and networking gear in much the same way – it can reconfigure the processors and drives of servers, arrange the networking gear to offer new services, virtualize a server into several virtual machines or combine several servers into a grid. Then it can change things up next month if desired. This sounds more like current assets than a bulldozer. But according to GAAP and the 800 year old double-entry math system we must use today, servers get placed in the same bucket as bulldozers.

My question is, how do hosting companies as an industry get together and establish some specific accounting standards that will allow our financial statements to truly reflect our business? Simply moving servers and networking gear to current assets would more accurately reflect how we use them in our business.

Am I off base in asking this? Hardly. The real estate investment business has been doing this for years. Traditional GAAP simply made no sense to their business, and they developed accounting standards that fairly represent that business. See this and note this quote from the Real Estate Information Standards, which is published by the National Council of Real Estate Investment Fiduciaries and is widely accepted among the real estate investment management industry and the firms that audit that industry:

"The development of the Market Value Accounting and Reporting Standards resulted primarily from the realization that standardization of meaningful financial reporting was necessary in order to allow real estate to become more acceptable as an institutional investment asset class. Accounting standards promulgated by authoritative accounting bodies exist for various real estate entities, including public real estate investment trusts and other public and private real estate entities that utilize historical cost accounting [i.e. GAAP – my comment]. The reporting requirements and information expectations of the institutional real estate investment community required the development of a market value-based financial reporting model for which no accounting standards published by authoritative accounting bodies presently exist. Accordingly, the lack of adequate authoritative guidance applicable to market value accounting for institutional real estate investment vehicles necessitated the need for these standards to be published."

Translated out of accountant-speak, this simply says that GAAP didn't fit their business and they applied common sense to the situation. I suggest that this young business of web hosting also needs some industry specific accounting standards to fairly report information about the health of its companies to its investors, and that these standards do not presently exist. Finally, if you've made it all the way through to here, you may order a server with free double RAM up to 2 GB by using the promo code "toothpaste&OJ" anytime over the next week. [subject to approval by Lance of course]


September 10, 2007

You Can’t Judge Health by Net Income Alone

In GAAP, net income is the bottom line. It's supposed to tell you if you're making money or losing money. But the amount on the bottom line is never equal to your bank balance and by itself, it's an inadequate measure of a hosting company's health.

For example, depreciation is subtracted before you arrive at net income. But depreciation is not cash going out the door. It represents the theoretical drop in book-value of something you own, such as servers. Ideally, the timing of depreciation should match the length of time you actually use something, so if you use a server for three years, its value would depreciate to zero over three years.

Problem is, from the years of hosting experience we have in this company, we know that servers bought in the early 2000's are still in use today. Many were depreciated over three years, but they're still generating sales revenue long after they've been depreciated to a value of $0. What this means is that net income from these servers was effectively UNDERSTATED during their first three years of use and that net income is currently OVERSTATED for the periods of use after their value has dropped to $0 on the books.

But since I have to judge a hosting company's health based on Net Income, here's how I do it. If net income is positive and it's greater than depreciation expense, one of two things is going on. Either they're knocking it out of the park in this asset-intensive business or they're not reinvesting enough to remain technologically relevant. If net income is positive but less than depreciation expense, they're likely healthy. If net income is negative and the absolute value of depreciation expense is greater than the absolute value of net income, the company could be fundamentally sound and worthy of receiving credit. Bankers will likely disagree with me, but my opinion here is hosting-business specific. Finally, if net income is negative and the absolute value of net income is greater than the absolute value of depreciation expense, then the company needs to adjust something to get healthy.

If you ask me, cash generation from operations is a much better indicator of the health of a hosting company. It ignores distortions like this mismatched depreciation. It will also tell you if the company generates enough cash to cover its debt service and/or to continue investing to stay technologically relevant.

Got that? If I haven't lost you already, I'll talk about how the normal Current Ratio calculation unfairly penalizes hosting companies next time.


September 6, 2007

Hosting and GAAP Accounting: Like Toothpaste and Orange Juice

Hosting and GAAP accounting go together like toothpaste and orange juice.

If you're confused go brush your teeth and drink a big glass of orange juice. I've held out as long as I can, but I just cannot restrain myself from a post or two about the hosting business and accounting. So if this will make your eyes roll back in your head, please stop reading now and click here before you keel over.

In many ways, good ole GAAP just doesn't treat the hosting business fairly. Relative to accounting, hosting is a new phenomenon with roots dating back only into the 1990s. This is ancient in Internet time but double-entry accounting dates back to the 12th century, and the first accounting textbook describing the double-entry system was penned by Luca Pacioli in 1494. The double-entry system was used because mathematicians denied the reality of negative numbers until the 16th century and the double-entry system was used as a workaround for the lack of negative numbers.

So, why must we account for paradigm-changing Internet businesses with an archaic 800 year old math system? It's a classic example of the old "square peg – round hole" cliché. Toothpaste and orange juice.

Applying this 800 year old system to the hosting business often paints a flawed picture of the financial position of a hosting company. And there are a lot of folks in the financial world that either can't understand this or don't want to understand this by thinking outside the normal accounting paradigm.

I'll blog about two examples of this: 1) Net Income and 2) Current Ratio. My next post will cover Net Income and we'll discuss Current Ratio thereafter.


August 27, 2007

Ditch Klunky, Hot-Running Servers for iPhones

Yes, servers shrink with size over time, but after reading this post about iPhones being web servers, I'm thinking about sending the following to David Letterman:

And now, the Top Ten Reasons Hosting Companies Should Replace Their Servers with iPhones

#10: At $500 a pop, they’re cheaper than new servers
#9: They come with more standard RAM than most servers
#8: They use less power than servers (just think how many you could cram on a rack!)
#7: They’ve got a multi-hour battery backup soldered right inside – so dispo your UPS’s also
#6: There’s no spinning disk drive to wear out
#5: You can ditch your bandwidth providers and leverage AT&T’s blistering fast EDGE network
#4: The operating system (Mac OS X) is pre-loaded – just rack it up and go
#3: Since you can’t crack it open, the expense of delivering custom builds is gone
#2: There’s a YouTube shortcut icon standard on every one
#1: They just look cool!



August 22, 2007

SaaS Who?

I'm always on the lookout for drivers of the hosting industry. One of those drivers is "Software as a Service" (SaaS). After seeing this article where Gartner thinks that the SaaS market will basically triple over the next four years, I wondered if the SaaS movement was affecting me yet.

I then realized that the SaaS market had grabbed me in one very important area – personal finance software. I had become disenchanted with my common desktop personal finance package – it had a lot of features that I didn't use and it didn't do some things that I need. So I began searching for options and discovered Mvelopes.

Mvelopes is a SaaS solution. Rather than buy the software, you subscribe to it over the Web. Anywhere you have an Internet connection you can access it, even from a mobile phone. It logs into the web sites of your bank accounts, investment accounts, retirement accounts, credit card accounts, etc., and brings your account info together in a convenient one-screen view. It will download info from accounts that your old desktop software can’t touch. And because it’s a subscription, you never have to worry about keeping up with new versions, patches, etc. Every time you log on you have the latest production software updates at your disposal.

It doesn't do asset allocation, portfolio analysis, technical analysis, stock screens or other fancy things that desktop personal finance software attempts to do. But it does planning and budgeting VERY well. This is what I need my personal financial software to do most of all, and my old software didn’t do this very well.

Probably the biggest hurdle in latching on to a SaaS solution like this is getting comfortable with placing the data security outside of your control since it doesn't reside on your local machine. But when you realize that the vendor hosts the data at a secure data center and has far better data security, physical security, and network security than your spare upstairs bedroom office, it is possible to make the move.

According to Gartner, software consumers will quickly realize the simplicity of subscribing to secure hosted solutions that are accessible from anywhere. Naturally, these rapidly growing SaaS solutions need a home. Consequently, we at SoftLayer would like to ask these SaaS providers such as Mvelopes, "Who does your hosting?"


July 20, 2007

Your Hosting Dollar

During some recent weekend R&R, my family and I saw a "human statue" street performer. He looked as if he'd been spray-painted gold – clothes, skin and all. He had a bucket out for "donations" and there was a healthy crowd watching. Parents would give dollar bills for their kids to put in the bucket. For each dollar, he'd do robotic movements and noise for 5 to 10 seconds and then return to statue status. After a few seconds, another dollar would go in the bucket and the cycle would repeat.

My son, a budding numbers-geek, said "Wow Dad, he makes pretty good money. I'll bet it’s $50 an hour." Being a full-fledged numbers geek, I said "By my calculations, it's more like $70 per hour".

This got me to thinking. What do we provide our customers for $1 of hosting fees? So I figured it out for our most popularly sold hosting offering. This is not $1 per line item below; it’s $1 for the whole package below.

  • 272,232,402,234,637 operations performed by the CPU at 50% utilization
  • 12 megabytes of RAM
  • 1.4 gigabytes of hard drive space
  • An Operating System to make it all happen
  • 45 seconds of technical support
  • 5,538,770,949,720,670,000,000,000 electrons (in the form of electricity)
  • 10,909 average sized packets of public transfer
  • Up to 37,973,200 average sized packets of private network transfer
  • All numbers are approximate. Nonetheless, be sure to make use of your hosting dollars here at SoftLayer!


July 9, 2007

Profit: A "Win-Win" Arrangement

Remember the "low-carb" diet craze a few years back? Some members of my family jumped on the bandwagon and I can remember seeing a lot of low-carb items in stores; low-carb milk, pasta, bread, chocolate, etc. Today you just don’t see as many of these products anymore. Look at the dates of the articles above and try finding some of the products in the links above – they’re long gone.

Why? Assuming these products really worked as advertised, when the low-carb craze was over, the cost of producing these products became higher than the revenue that the market was willing to pay for them. Maybe the market rejected them because they didn’t work. Whatever the case, mathematically, when costs are higher than revenue, there is no profit. Consequently, companies stopped offering these money-losing products. No profit is a "lose-lose" situation. Neither the companies nor the consumers who want the discontinued products benefit when there’s no profit.

The same goes for the hosting industry. If the cost of providing hardware, software, power, cooling, and bandwidth ever rises higher than what the market demand will pay, this offering will exit the marketplace. Personally I don’t think that will ever happen. Because there is an opportunity for profit in the hosting business, we and other providers will continue to inject our offerings into the marketplace. And due to the cost of these offerings, we won’t be offering dozens of processing cores, unlimited RAM, unlimited bandwidth and multiple terabytes of storage capacity for ten bucks a month.

Thankfully, SoftLayer doesn’t have to deliver all of that to achieve a top notch customer experience (as of yet anyway). But simply providing the list above is only part of the equation. As I mentioned in my last post, treating your customers "right" and building long-term relationships is critical to maximizing profit. Therefore, we do our best to price our offerings at value points that make both our customers and our investors happy. The resulting profit ensures that we continue in business and that we keep our server fleet refreshed. Profit keeps us around and motivates us to provide our customers with an excellent customer experience.

Thus, for SoftLayer and our customers, profit is a "win-win" situation.


June 29, 2007

Business Ethics Simplified

In this day and age of Sarbanes-Oxley internal controls, SAS 70 certifications, and myriad other regulatory, compliance, and audit issues that I won't get into , business ethics might seem to be a lengthy and complex topic.

In reality, it isn't. Back in the dark ages when I strolled the halls of SMU, a crusty Econ 101 professor named Jack Stieber proclaimed that there is only one ethical mandate in business: "Within the bounds of the law, maximize profit." There are no more ethical rules necessary to follow in business.

I have heard others phrase a similar thought as "maximizing shareholder value". I disagree with that approach because there are things that management can do to influence the stock price that aren't necessarily tied to maximizing profit. Basically, if you can maximize profit, the stock price will take care of itself.

In response to Prof Strieber's proclamation, there were a few students who responded, "But sir, what about ?" and Prof. Stieber shot them all down. Here is one of the more interesting objections:

"But sir, what about a business owner who hikes the price of bottled water to a ridiculous level in a disaster-stricken area that has lost its water supply? Are you saying he's being ethical by maximizing his profit from price gouging?" Prof. Stieber responded something like this:

Assuming that his pricing policy is legal, he's still being unethical because he's actually not maximizing his profit. Sure, he may reap a short-term gain but when the water supply is back on, those forced to buy his extortion-priced water will take their business elsewhere. So in the long term, he hasn't maximized his profit and thus has behaved unethically. An ethical decision during that time might have been to keep selling water at the pre-disaster price or maybe even donating some to build goodwill among his customer base. This could have cemented a long term relationship with the customers who would provide repeat business again and again and thus maximize his profit over time.

That being said, when a business maximizes it's profit within the bounds of the law, it's a "win-win" for the customers, stakeholders, and shareholders. In my next post, I'll explain how SoftLayer earning profit is a win-win for both the customers and the company.


June 4, 2007

Why Finance Guys Don't Blog

Q4UY don’t finance guys blog much? If j00 post “IAAA” and talk of KPIs, EVA, and other TLAs, readers think listening to this llama is a CWOT and say “CYAL8R”. CMIIW but hosting demand r0x0rs. The SMB market sk00lz all else but there are other factors. I’ll mention just a couple here:

I’ll call one the “middle school” factor. I have a 13 year-old boy. He and his classmates are absolutely addicted to Internet chatting. He’ll open six or more windows at once and at least four of them are girls who are also chatting with IDK, their BFFs AFAIK. They will ROTFLOL for hours even if OMG, PAW. It’s NBD to them.

I doubt that as these kids grow up they’ll give up the chat habit, and the n00bs that come along will only add to the ranks. Thus, another driver of internet fundamentals grows seemingly forever and demands more servers to relay the ever growing messages.

I’ll call the other factor the “mullet factor”. I knew our CEO back in the 80’s and he sported quite the mullet, I can assure you (see image to the left for proof).

Punch in the word “mullet” into Google and in .05 seconds you’ll get links to about 3.8 million web sites somehow related to mullets. w00t! A few are related to the fish, but most have to do with the hairstyle. YKW, these websites have to live on a server somewhere. Strange websites like this only seem to proliferate over time. AWHFY?


Subscribe to Author Archive: %