It's Rishi

The price of oil is creeping up fast. Right now as I write this post, oil has just risen above $93, a 50% increase over the average price of oil in 2006 (about $60 according to this data). While the current crude oil prices are in absolute dollar figures a record high, on an inflation-adjusted basis, we are not quite at the all-time record level of ~$98 set in 1980; source).

Without getting into the details of why oil prices are rising and if the trend will continue, let’s discuss a certain key question. A co-worker of mine recently asked me if I had changed my driving habits as a result of the increased gas prices in the past year. That was an easy answer for me: “No”. Then he asked the more interesting question: “At what price will you change? $4? $5? $10?”. $10 gas may sound ludicrous, but it’s really not that farfetched. After all, gas prices in most of Europe is at least twice what it is here. In fact, gas is approaching $9/gallon in the UK! (source)

So, what is my magic number? It’s hard to throw out a number but I do know that it’s a high number. My weekly routine does generally involve a fair amount of driving, and there’s few to zero suitable public transportation solutions. Let’s say I drive 1k miles per month and my car averages 25mpg. That’s 40 gallons a month. At current gas prices that’s about $150. At $10/gallon, that’s $400. Annualized, that’s $5,000, or about $8,000 on a pre income tax basis! If gas prices went extremely high, I would undoubtedly need to change my living requirements. Specifically the route from home to the office would either need to involve a short drive or primarily public transport. My driving habits and the fuel economy of my car are very average. With that in mind, let’s remember that the US median household income is not much more than $40,000. About 15% of your post-tax income dedicated just to gas is an impossibly high expense.

The truly scary aspect of rising oil prices is its compounded effect on the cost of living. It’s no secret that higher oil prices correlates with inflation. Not only does high gas prices increase the wallet burden at the gas pump, it also affects the price of many common consumer items because the costs associated with transporting those items rises.

It is often said that Americans view cheap oil as a God-given right. Whether that’s true or not (I tend to agree that it is true), we may soon find that right revoked for good. And at the current rate, it may happen a lot sooner than anyone had predicted. How that affects the landscape of this nation in the decades to come will be absolutely fascinating to watch.

I was reading this post by Paul Kedrosky talking about how Amazon’s gross profit margins have been consistently declining since 2001, which sort of contradicts the company’s claim, since its inception, that massive scale will make the business more profitable.

I realized that I wasn’t exactly sure of the the definition of gross profit margin versus operating profit margin. I googled it and found this Income Statement Analysis guide on About.com. Before I knew it, I had spent a good half an hour going through most of it. I found it to be a very practical crash course.

Some interesting things I learned:

1. For asset-intensive businesses, how the company accounts for depreciation is a really important thing to consider. Depreciation methods, rates, and salvage values can make big differences in a company’s bottom-line. More specifically, they can be easily played with to make the company’s earnings look better.
2. EBITDA is BS. Creditors demand interest payments, the IRS demands tax payments, and depreciation and amortization are absolutely real costs of earnings. EBITDA is sexy but really misleading.
3. Comparing operating margin versus gross margin across many competitors in the same industry can tell you a lot about each company’s business models. Like this post that talks about Tim Hortons vs. Starbucks.
4. If a company owns a minority stake (< 20%) of another business, it only reports the actual cash (dividends) it receives and NOT the % portion of earnings that the owning company is entitled to. This profit that the owning company is entitled to but doesn't report is referred to as "look-through" earnings. In one famous example, Berkshire-Hathway, actually had it's look-through earnings exceed it's own earnings, so Berkshire's actual earnings report was a horrible understatement of the company's performance.
5. “Record earnings” don’t necessarily mean jack. ROE (return on equity) is the number to look at. After all, a company which generates the exact same operating income every year, but sticks all its earnings in a 5% interest bank account, will consistently have record earnings simply because of the increased interest income on increased equity. Shareholders can put their own money in the bank and earn that interest. They don’t need a company to do it for them.

Anyways, I would highly recommend you at least browse through this guide if you aren’t already familiar with these concepts. I was somewhat familiar with many of these concepts, but this guide really tied everything together nicely. Thanks Joshua Kennon!

While I do actively manage my personal investment portfolio, I still consider myself to be a fairly green investor. I own just a few individual stocks that I’ve held long positions in. Most of my money is invested in about 15 mutual funds. One aspect of trading that I’ve always been curious about, but never actually tried was options trading. I understand the mechanics of options but I never understood if, when and how I should be utilizing this form of trading.

A college buddy of mine, Kevin Chou, recently launched FinanceHuddle where he discusses his investment strategy and his own trading activities. I asked Kevin if he could summarize options trading and contrast it with buying/shorting stocks. He graciously did so in his latest post, Using Options – The Good, the Bad, and the Ugly. I had to read it a couple times over to fully digest what he wrote, but having done so I feel like I have a much better understanding of options trading. Thanks Kevin!

My friend Rob Poitras pointed out to me that YouSendIt, one of the most popular free file hosting services, serves up a whopping 44TB of data every single day. He mentioned that as far as he knows, they are not profitable either. So, I was curious about how sound their business model is. I have to think that it is because just in the past year, there has been many other essentially identical services that have launched.

Unfortunately, I couldn’t dig up any hard numbers so we’ll have to base this analysis off of fairly rough details. The facts from YouSendIt.com claim 44TB of data daily and “over 8 million unique visitors” monthly.

Costs
It doesn’t take a genius to realize that by far the primary cost of doing business for a file hosting service is servers and bandwidth. I’m guessing the latter far outweighs the former. Not only are they serving an insane amount of data, but also for a file hosting service to compete, it needs to offer fast download speeds to users.

44TB of data flows daily and there is 86,400 secs in a day. At any given time, there is an average of 0.5GB/sec (which is 500MB/sec or 4000Mb/sec) being transferred. I really wish I had a good number of how much it costs to serve data on like a per GB basis. I can imagine some ways I could go about calculating it but without accurate starting data, it’s probably not worth the time. I saw some numbers like 1U rack space + 100Mbps unmetered for $2500/month. Let’s just say that you bought 40 of these. Each 1U file server maxes out its 100Mbps connection. Let’s say each file server is leased and runs about $500/month. So total cost for all this is 40 x ($2500+$500) = $120k/month. Obviously this is a very crudely made number and I’m sure that the scale that YouSendIt is doing things, it should be cheaper. However, there is a significant operational cost to keep such a large service like this running and there is some development cost for various software and tools and such. Let’s say total for everything costs $175k/month. (I still did not include general personnel and administrative costs for YouSendIt)

(Advertising) Revenue
YouSendIt makes money via advertising. However, because this advertising space is not very targeted (really the only way to target is the name of the file which is not necessarily meaningful at all), I just can’t see it having as high a CTR (click-thru-ratio) as a contextual systems like AdSense or keyword systems like AdWords. YouSendIt has two types of ads on their site, image/banner ads and text ads. My *guess* is that the image/banner ads are sold on an impression basis not CPC. I think that is typically the case with image ads because they are so imposing on a page, that it’s hard for the user not to notice it. The question is how expensive? Since it’s just not that targeted and users don’t really spend that much time viewing the page (just click download link and leave), I could see it going for $5 CPM (cost-per-thousand-impressions) which is $0.005/impression. For text ads (looks similar to AdSense text ads but it’s managed by a company called BidClix), depending on the type of business the advertiser is, YouSendIt charges a minimum of either $0.25 or $0.50 for each click. Looking at the current bid amounts which are a few pennies above the minimums, let’s assume an average of $0.40/click.

So to calculate monthly ad revenue, let’’s refer to the 8 million unique visitors per month. Let’s say that results in 10 million file-download-page views (the pages that show ads) per month (average of 1.25 page views per unique user per month). Each page view includes 2 of the banner/image ads so that’s 20 million impressions. At our figure of $0.005/impression, that’s $100k. For the text ads, since again they are not very targeted, a CTR of say 1% seems reasonable. 3 text ads are displayed at a time. So total clicks on average per page is 0.01 x 3 = 0.03. Multiplied by 10 million page views that’s 300k clicks per month. At an average of $0.40 per click, that’s a total revenue of $120k for click revenue. Total ad revenue estimate is $100k + $120k = $220k.

Final Analysis
So my estimate cost of operating this service is $175k and estimate of ad revenue is $220k on a monthly basis. So we got a profit right?! Well, as I mentioned earlier, I didn’t include administrative costs. Let’s just assume that’s a paltry $25k/month which is mainly salary for the key administrative individuals. Luckily the service sort of markets itself virally so I’m guessing marketing costs are very low. Anyways, so *total* montly cost is $200k and revenue is $220k. $20k/month profit..we’re in the black!

Again, this is an extremely rough conclusion. I am curious how far off I am though. What does interest me is that I think some of these file sharing sites received venture investment. I assume this means that these sites plan to leverage their traffic or content (somehow) into a more profitable business. The content seems like a tough play since it’s private and also a lot of it is probably (illegal) copyrighted material. Without such a long-term vision though, sites like these seem to me like a commodity. Even if I were to double my estimate for CPM to $10, the company still is making about a $1MM profit annually. The problem is growing that. Maybe a better operating margin can be squeezed out by improving efficiency or somehow offering more targeted advertising, but still it doesn’t seem like it would offer the growth potential that venture investors would want to see.

Anyways, I’d love to hear some comments on this analysis. If any YouSendIt guys read this, wanna tell me if my analysis is even remotely accurate? =)