John Swift's  Instablog

Summary
Edgar Online (EDGR) is a small cap technology company that processes public filings from the SEC's EDGAR database and makes them available in various formats to investors. The company has never had a profitable year in its 14 year history (founded 1995, public in 1999). Nevertheless, it trades at an eye-popping 9x book and 40x EBITDA. Hope springs eternal, but I believe EDGR is overvalued.

 

The bull story

EDGR’s nosebleed valuation is largely credited to the company’s leadership in an emerging technology called XBRL (eXtensible Business Reporting Language).  XBRL is a markup language that promises to transform SEC filings from documents to standardized data that can be processed more efficiently by computer programs. Spreadsheets can automatically pull in every number from the filing and “understand” what it means. Your integrated 3-statement spreadsheet model will automatically download and plug in numbers from the latest quarter. It will work not just for the income statement or the balance sheet, but for every number in every footnote. That means you can get an automatic comparison of the company’s backlog or number of customers (for instance) quarter over quarter, or year over year. You can do it for one company or the whole industry, all before reading the first line from the 10K. The potential applications are almost limitless…

 

To make XBRL work, someone is going to have to mark up all those 10Ks, i.e. pull out the numbers and classify them in a standard way.  This seems like a lot of work. Why would any company want to make this investment? Because it’s been mandated by the SEC.  The 500 largest US companies were covered by the mandate in 2009. 1,800 companies are covered in 2010 and all 10,300 reporting entities in 2011. 

..And that’s just Qs and Ks in the U.S. If you count filings for debt and mutual funds, and then count filings outside the U.S. you can get some truly large numbers. EDGR’s latest investor presentation puts the addressable market at about $2B (although that estimate tellingly omits the year in which they believe the market will get to that size). On top of that, there’s the market for translating, data-basing and reselling old filings in XBRL format - there’s another $5B, according to the company.

 

EDGR’s first XBRL product debuted in 2005 and it currently has 36% of the market for XBRL-based SEC filings (Q3’09 thru 8/3), including high-profile clients like Google, Yahoo and Intel. If EDGR can maintain its share as the market grows, we would be looking at a $1B company. Much better than the $19m in revenue from the past 12 months.

 

 

EDGR’s business

Let’s step back from the hype: EDGR makes money three different ways:

1. It sells Subscriptions to web-based tools that allow users to analyze SEC filings.
2. It sells Data, i.e. SEC filings (raw, parsed and formatted) directly to the user’s systems.
3. It helps US companies prepare XBRL filings for the SEC, as described above.

 

XBRL is the smallest segment at 28% of 2009 Q3 revenue. It’s growing fast but Subscriptions are shrinking almost as fast, so total revenue is growing slowly:

 

	2008 Q3	2009 Q3
Subscriptions	2,127	1,585
Data and solutions	2,110	2,179
XBRL filings	464	1,478
Total revenue	4,701	5,242

 

I believe the reason Subscriptions are shrinking also has to do with XBRL: EDGR’s original value proposition was to provide a superior user interface to that available on the SEC’s (free) EDGAR site. You could go to Edgar Online and download financial statements in spreadsheet-format, whereas on the SEC site you could only get the raw filing. Now, you can download the spreadsheet format from the SEC site for the largest 500 companies who are filing in XBRL, and soon for the rest. I’ve tried that interface – it’s not half bad, and it’s going to get better. Also, thanks to XBRL, there are now various tools available (some free and open source, like the one from the SEC) that close enough to the Subscriptions offering where it counts. They are not as polished, but they are getting there. In other words, XBRL is undermining the value proposition of the Subscriptions business.

 

Competition

EDGR operates in a highly competitive environment. The Subscriptions and Data businesses compete with much larger and better capitalized companies, such as Thomson Reuters, Bloomberg, the big 3 rating agencies and even Factset.

 

The fledgling XBRL filing business (<1 year old) is also highly fragmented, contrary to the company’s PR. EDGR is in the top spot with 36% of filings in 2009 Q3 (according to the company). Bowne is in the second spot with 29% and self-filed returns make up 19% of the market, the third spot. Then there is the hodge-podge of small software vendors offering XBRL tools. And, there’s BusinessWire, which is owned by Berkshire Hathaway and has just introduced an XBRL tool called CoreFiling.

 

Strategic problems with the bull story

The biggest problem with the story is strategic: It is just not possible to create the highly profitable software business that is implied by EDGR’s valuation based on a free, open, ubiquitous standard. Netscape and Red Hat are two cases in point although the dot com graveyard is littered with others. This is the biggest reason why EDGR hasn’t been able to turn a profit yet.

 

EDGR simply does not add enough value with its services. The technology is basically a parser that converts text documents in a semi-standard format into financial data. There is no value-added content or intellectual property beyond this. As a former programmer and software entrepreneur, I can tell you creating this type of parser is far from rocket science. Yes, there are a lot of cases to work through, but it is just not that hard to do.  We can see just how little EDGR spent on its technology from the balance sheet: Capitalized software in 2009 Q3 totaled $250k plus $1.4m for “internal use.” In other words, you could replicate EDGR’s software, both internal and client-facing, for less than $1.7m. In reality, you could probably replicate EDGR’s software for considerably less, because when an open standard like XBRL matures, supporting vendors and tools develop (fashionably called an “ecosystem”) and it becomes much cheaper to create the same output XBRL document. This puts incumbents like EDGR at a disadvantage relative to newer vendors with cheaper technology. This type of development is an inevitable part of standardization. We can see it in action in almost every area of the software business. Making XBRL implementations cheap and widely available was also one of the SEC’s stated policy goals when it mandated XBRL. They will absolutely not allow one company to dominate.

 

My hunch is that EDGR has some automation advantages over its competitors which help explain its market leading position (by 5% points), but these advantages are very fleeting, not sustainable.

 

Catalyst: Deal with RR Donnelley

All of EDGR’s XBRL filings business comes through its partnership with RR Donnelley. Signed in 2008, this partnership makes EDGR the exclusive XBRL implementor for RR Donnelley clients for a period of three years. Prices for year one were set up front. Prices for years two and three were supposed to be negotiated by Oct 2009 and 2010 respectively. What interesting is that the Oct 2009 negotiation is behind schedule – year two prices have not been set yet. I think this is because RR Donnelley is realizing it has many alternatives to EDGR, and therefore deserves a much larger share of the revenue than it took in year one.

 

I believe this partnership is EDGR’s biggest asset as well as the first catalyst to bring its valuation down to earth. When the negotiation is done, I believe we will find EDGR needs RRD a lot more than the other way around.

 

 

Disclosure: No positions

TGC Industries

TGC Industries, Inc. (NasdaqGS: TGE) is a small-cap seismic survey company that helps oil and gas companies find oil and gas.  Over the past 5 years, TGE performed spectacularly well: it grew its revenue >12-fold, its fully-diluted EPS > 18-fold and its free cash flow >32-fold. However, its sales have stumbled badly in 2009Q2 in a sector-wide rout. As a result, it is available for what we believe is an attractive price of 3.8x EBIT and 2.8x FCF.

Seismic data acquisition industry

Seismic data acquisition companies are hired by oil and gas exploration companies to blast sound (and sometimes dynamite) into the ground (or underwater), capture the echoes that come back and formulate an underground map of the Earth’s strata which can help pinpoint the locations where oil and gas are most likely to be found. This in turn helps optimize drilling activity. It also helps oil and gas companies “manage” more mature fields by delineating exactly where the oil and gas pockets end, and how much is left.

Seismic surveys are performed by “crews.” Each crew numbers 35-45 members, and a lot of specialized machinery. One crew costs around $7m to operate, all in. TGE discloses the number of crews it operates on a quarterly basis.

Seismic survey industry growth is primarily driven by oil and gas prices. As these prices rise, more exploration activity becomes economically feasible, and more seismic data acquisition projects are commissioned. As the price of oil rose from around $40/barrel in 2004 to about $90/barrel in 2008, the number of seismic crews operating in the US (onshore) rose from 50 to 73. However, as oil dropped back to $40-45/barrel, the number of seismic crews dropped to 60.   So overall, the number of crews has declined more slowly than the price of oil, and is at higher levels now than it was the last time the oil prices were at current levels. We believe this indicates that seismic data surveys have become more of a competitive necessity and demand for these services is likely to grow faster than the broader energy industry.

At first blush, seismic sounds like a highly volatile, high-tech industry where an upstart can easily displace incumbents through superior technology. TGE and its competitors boast about their “state of the art equipment” on websites and in SEC filings. However, we have reason to believe the pace of innovation is not so rapid, and barriers to entry are higher than they seem:

• Market concentration: The top four public competitors deploy about 50% of the crews working in US onshore. The competitors are TGE, Dawson Geophysical (DWSN), Geo Kinetics (GOK) and CGG-Veritas (CGV).
• Market share stability: Total market share for the same top four competitors has held remarkably steady at ~50% since at least 2005. Each of these competitors have been around for a lot longer.
• Slow technology adoption curve: Seismic data gathering is done using several methods. The oldest is “2D.” The most currently popular method is “3D” and the newest is “4D.” We won’t go into technical detail on the exact differences between these methods, but suffice it to say the newer methods are more technically sophisticated and more expensive. If seismic were a highly innovative industry in high flux, one would expect that newer and newer methods to be introduced and adopted in quick succession, kind of like PC chips or hard drives. However, adoption of the “3D” method went from 88% of US onshore crews in 2000 to 96% in 2009, giving us +8% adoption in 10 years – hardly Moore’s law. To be sure, there are incremental improvements in the technology deployed, but we are skeptical of the “permanent revolution” PR being promoted by the industry.
• Senior-level sales: According to TGE, most sales depend on interaction with customers at the executive level, and are “based principally upon professional relationships developed over a number of years.”  This could help explain why the cast of players does not seem to change much from one year to the next.

Why is TGE different?

TGE has been in business since 1986, when it was spun off from another Texas-based seismic company. TGE distinguishes itself from its top three public competitors (DWSN, GOK, CGV) in a couple of strategic ways:

a)      Market focus: TGE focuses exclusively on the U.S. exploration market. Furthermore, most of its work is done onshore. A small minority is done on land-to-water. The three competitors have a much more diffuse geographic focus, and do both ground and water operations.

b)      Standard equipment: TGE standardizes the equipment among its crews to ARAM Aries. We believe this is a significant and under-appreciated benefit: It makes crews and hardware more fungible, and improves the logistics: Every crew is trained on every piece of hardware. It also improves TGE’s purchasing power with its vendor.  TGE is smaller than its four competitors, so it’s easier for TGE to maintain a standard hardware environment.

c)       High insider ownership: Officers and directors own 28.7% of TGE shares outstanding, a much higher percentage than at competitors. We believe this is primarily a positive, in that it provides a lot of shareholder/owner alignment. This alignment was evident in TGE’s aggressive cost cutting in response to slowing demand in 2009 (more on this below).

As a result of these differences, TGE has achieved higher returns on capital (ROIC) and faster revenue growth than competitors over the past 5 years:

	5yr Avg ROIC	Revenue Growth, 2009 vs. 2003
CGV	28%	5.49x
DWSN	21%	5.46x
GOK	3%	14.79x
TGE	36%	12.27x

 

Only GOK grew faster than TGE, but this growth came at the cost of much lower profitability, and was probably value-destroying: GOK had negative free cash flow of over $40m in the five year period.

TGE’s revenue growth happened because of market share gains as well as productivity increases:

• TGE’s share of US onshore seismic crews from 6% in 2004 to 12% in 2008, and
• Annual revenue per crew rose from $6.7m to $9.6m in the same timeframe. We believe this reflects price increases as well as incremental increases in the sophistication of the equipment and services rendered.

Why is TGE Cheap?

Considering TGE’s growth and profitability, we believe it is cheap on a couple of metrics, including FCF/EV of 36% and EBIT/EV of 26%, calculation below:

Shares Outstanding (m)	18.00
Price per share, 10/16/09	$4.82
Market Cap ($m)	86.76
Debt ($m)	14.90
Cash & ST Investments ($m)	30.60
Enterprise Value ($m)	71.06
FCF, TTM 6/09 ($m)	25.90
EBIT, TTM 6/09 ($m)	18.60
FCF/EV Yield	36%
EBIT/EV Yield	26%

 

Why is TGE so cheap? Perhaps the biggest reason is the sharp decline in its number of crews: From 2006 through 2009 Q2, TGE operated at a steady 8-9 crew level. At the end of 2009 Q2, it downsized to four crews. This was a pretty drastic drop. In the latest analyst call, CEO Whitener said “we have been experiencing weaker demand... We are also see continued price pressure and therefore our margins have deteriorated and are off from the normal levels. From my perspective this is the most severe downturn occurring in the shortest period of time in all my years of being in the business.”

This downsizing happened too late to affect revenue in 2009 Q2 much, but it will certainly take away from 2009 Q3 and Q4 revenue. The question is: how bad will it get?

Downside scenario

Let’s assume TGE continues operating with four crews instead of eight, and does not grow further, ever. Considering TGE’s past record of profitable growth, this would be a very conservative scenario. TGE’s annualized revenue per crew for Q1 and Q2 2009 was about $14.7m. That’s if we assume conservatively that it operated 8 crews throughout the two quarters, and only downsized to four crews on first day of 2009Q3. From there, we can construct a pro-forma earnings-power estimate. In this scenario, EV is 6.8x EBIT: Not a screaming bargain but not very expensive either.

Pro-Forma Income w/4 Crews
Annualized Sales / Crew	14.7	Based on Q1, Q2 2009
Crews	4	End of Q2 2009
Total Revenue	58.6
EBIT Margin	18%	2004-08, Avg
Steady-state EBIT	10.46
Current Enterprise Value	$     71.06
EV/Steady-state EBIT	6.8	x

 

We think it’s interesting to note that TGE was very aggressive in cutting costs (crews) when it saw its demand dropping – more so than competitors – and was therefore successful in staying profitable and even generating peak FCF through the down-cycle so far.

In reality, as we’ve seen above, TGE has had the most profitable growth of its competitor universe. It is also expected to grow the most in the future: analysts forecast TGE’s EPS to grow 41% over the next 5 years versus 20% for GOK and 10% for DWSN.

We do not believe the drop in crew capacity is due to a unique aspect of TGE because it can be observed industry-wide: U.S. onshore seismic crews dropped from 72 in March 2009 to 60 in June 2009, a 17% drop.

Upside Scenario

The upside scenario is that TGE returns to the 8 crew level (say, by 2010 Q3) at which it’s operated over the past few years. TGE is currently trading at 2.75x FCF whereas its competitors are at an average of 6.16x FCF. If we apply the competitors’ multiple, TGE would be worth about $10.80/share.

There are a couple of other recent developments that could provide additional upside, although they are not counted in the valuation:

• On Oct 16th, TGE bought Eagle Canada, a small seismic data and survey provider based in Calgary, out of bankruptcy court. TGE paid $10.3m in cash. Eagle Canada had revenues of $26.5m in 2008, so TGE paid about 39% of 2008 sales. Given TGE’s valuation, this should be accretive right away. I think this acquisition also illustrates the types of opportunities available to players with the strongest balance sheets, like TGE.
• TGE has begun assembling a “data bank” of gravity and magnetic data covering many of the major oil and natural gas producing areas in the US. This can potentially lead into a new business model where companies purchase the data rather than commissioning a new survey each time. The data bank does not bring in significant revenue yet, but we think it could be quite interesting if successful.

Catalyst

• TGE returns from 4-crew level to something closer to its long term steady-state of 8 crews
• Credit flows more freely to exploration companies, boosting their CapEx, which in turns drives more Seismic business in the US
• Eagle Canada sales and crews show up in TGE’s financials in 2009 Q3, Q4

Disclosure: No position in TGE

SumZero Writeup: 3x ETFs

This is not an original writeup, but I like the idea very much and I think SA readers would benefit from seeing it. And, as Picasso said "Good artists copy; great artists steal."

The idea is simple:
Take a pair of 3x and -3x ETFs, such as Direxion Daily Financial Bull/Bear 3X Shares (FAS & FAZ), and short them both. Both securities in the pair should lose value in the long term, and if you rebalance the position, you can maintain it to be pretty close to delta-neutral. The same idea should work with any pair of 3x ETFs, the more volatile the better.

Why would this work?
A 3x bull ETF has a daily exposure profile similar to 100% equity and 200% debt (the actual exposure is achieved using derivatives, but the effect is the same). This ETF tries to exactly triple the daily return of the underlying stocks it tracks. To do that, it tries to maintain this exact leverage ratio (1:2) throughout the day. What happens if the basket of underlying stocks it tracks drops during the day, say by 1%? The value of the equity drops to 99% and the value of the debt remains unchanged at 200%. The ratio is now 0.99:2, higher than the target. So the ETF will sell assets (stocks) and pay off debt (unwind derivative positions) until the ratio is back at 1:2. The ETF *has* to do this to keep its intraday tracking error to a minimum. The effect is locking in losses when stocks go down.

The opposite happens when the underlying stocks go up 1% intraday: The leverage ratio drops too low, the 3x ETF has to buy more assets (stocks) and take on more debt (increase derivative exposure). In other words, this bull ETF buys when stocks go up and sells when they go down. The bear ETF buys when stocks go down and sells when they go up. In both cases, it's exactly the opposite of what a value investor would try to do.

It gets worse, because the ETF may need to make multiple position adjustments throughout the day, depending on the volatility of the underlying. This increases transaction costs and taxable short term gains. That's why 3x ETFs have some of the highest expense ratios among ETFs.
 
We can see the value destruction in the price history of a FAS and FAZ:
On a 1-day scale, things look OK. FAS and FAZ are near-perfect mirror images:
finance.yahoo.com/charts?s=FAS#chart10:s...;range=1d;compare=faz;indicator=volume;charttype=line;crosshair=on;

On a 5-day scale, you can start to see a bit of degradation, note that on day 5, FAS is down a bit more than FAZ is up (10% vs. 9%)
finance.yahoo.com/charts?s=FAS#chart11:s...;range=5d;compare=faz;indicator=volume;charttype=line;crosshair=on;ohlcvalues=0;logscale=off;source=undefined

If we extend the timescale to about a year, giving the market more chances to change direction on a day-to-day basis, the funds inexorably lose more and more value. Bad if you're long. Good if you're short.
finance.yahoo.com/charts?s=FAS#chart8:sy...;range=1y;compare=faz;indicator=volume;charttype=line;crosshair=on;ohlcvalues=0;logscale=off;source=undefined


Multi-day example:
The example above discusses how an ETF maintains performance intraday. Let's look at what happens day-to-day: Let’s say the index of the underlying financial equities tracked by FAS starts at $100 on day 1, goes up 5% to $105 on day 2, then drops 4.76% on day 3 to end back at $100. So the underlying securities broke even. How did FAS do? On day 1, FAS went up 15% and ended at $115. Great. On day 2, it dropped 14.29% and ended at – wait for it - $98.57. Why not $100? Because the loss was applied against a higher starting balance. The reverse happens on the bear fund.

You can play with this scenario a bit and try different percentage gains and losses. The bottom line is, each day the market changes direction the securities lose money. The more volatility the market experiences, the greater the value loss experienced by the securities.

References
- If this sounds interesting, I would encourage you to read the more detailed writeup by casper719 on valueinvestorsclub.com (idea # 14325)

- How 3x ETFs behave in trending and non-trending markets: etfdb.com/2009/the-truth-about-3x-etfs-a.../

- More in-depth explanation of how (2x) leverage ETFs work. 3x ETFs work the same way. Not in particular the "constant leverage" section. seekingalpha.com/article/35789-the-case-...

- List of 3x ETFs: etf.stock-encyclopedia.com/category/trip...

- More 3x ETFs coming from ProShares: www.indexuniverse.com/sections/newsinfoc...



Disclosure: No position in FAS or FAZ