In recent years, a number of entities have recognized that Canadian-based seismic data provider Pulse Data Inc. (OTCQX:PLSDF) [TSE: PSD] is extremely undervalued. A variety of companies have attempted to purchase PSD for dirt cheap valuations and fortunately for PSD shareholders, have been rebuffed by the Company’s board of directors. Seitel/ValueAct (“SVA”) is the latest group to approach PSD although its recent bid for $3.10 per share is its second attempt to purchase the Company in the past two years. With approximately 13.5% of PSD shares, SVA is attempting to purchase the Company for just 5.2x EV/2007 EBITDA (Kinnaras estimates).
SVA stands to earn an incredible rate of return if it succeeds in acquiring PSD for this cheap valuation and assuming its cash flows. While I admire Jeff Ubben’s team and hope to emulate ValueAct’s success myself, I personally owned Seitel when it was purchased by ValueAct at a fairly low price due to ValueAct owning over 50% of the company when attempting the buyout and felt ValueAct got an incredible deal. I owned PSD personally during that time as well and upon launching my fund, bought shares for the fund due to the incredible value in the Company. Running a summary returns analysis of a combined SVA/PSD clearly illustrates that SVA can afford to pay much more than $3.10 per share to purchase PSD.
SVA and PSD are involved in the seismic data industry. SVA owns the largest set of North American seismic data while PSD commands the second largest set of seismic data in Canada. Seismic data is critical to oil and gas exploration, representing about 5% of the total exploration costs but accounting for 60% of the decision making process. These companies generate revenues either by licensing their data sets out to exploration teams or by conducting participation surveys.
PSD has focused mainly on data licensing due to its commitment to cash flow production. Licensing costs for 2D and 3D data run from $300-$20k per km and $6k-$7k per sq km, respectively. In contrast, it costs $7k-$90k per km and $30k-$100k per sq km to shoot new 2D and 3D data respectively. Participation surveys, although partially funded by clients, offer challenges due to the volatile costs of work crews and weather conditions. As such, PSD is very selective when engaging in participation surveys although the Company does recognize the importance of conducting surveys for clients.
To understand why SVA’s offer severely undervalues PSD, it’s important to understand that both business models are very lean, requiring little overhead, and produce cash margins of approximately 80%. As such, integration is fairly easy as the data sets are quickly assumed and overhead/corporate costs can be eliminated. Table I presents a summary income statement for Seitel and PSD.
[click to enlarge all images]TABLE I: SEITEL SUMMARY INCOME STATEMENT (USD MM)
The Pro Forma Annualized column is utilized to present the true operating performance of SVA on an annualized basis for 2007, excluding its one-time merger costs. Q3 is generally very volatile but in recent years has performed well while Q4 is typically the strongest. This mirrors the first half of the year where Q1 is generally as strong as Q4 while Q2 is usually the weakest of the year. As a result, it’s not unreasonable to simply double first half operating performance to annualize operating performance. Table II next lays out a combined income statement for PSD and Seitel. PSD’s 2007 estimates are internal and are in USD to match Seitel’s reporting currency at an exchange rate of CAD1.06/USD 1. The synergies eliminate PSD’s costs, an assumption that is entirely plausible because the data can simply be transferred to Seitel and be sold through Seitel’s internal sales force. Further, PSD is already an extremely lean company with just 25 employees.
TABLE II: SEITEL/PULSE SUMMARY INCOME STATEMENT COMBINATION (USD MM)
As Table II illustrates, the combined company will maintain 80% EBITDA margins, as good as it gets for any leveraged buyout model. At SVA’s offer price of $3.10, SVA needs just $146M to purchase the 47.1M shares it does not currently own. SVA current’s capitalization is $25MM in cash and $403M in debt, resulting in Net Debt/EBITDA of well under 4.0x. While this may seem high to some, this is far below average for today’s leveraged buyouts, even with the current credit turmoil. I would expect that SVA will present pro forma figures similar to those in Table II to its lenders, demonstrating that the combined EBITDA would be roughly equal to the funding required and easily borrow the $146M to fund its purchase of PSD. The additional funding to acquire PSD and refinance its existing debt and assume its cash would result in a cash balance of $34M and $584M in debt against $143M in combined EBITDA or just 3.8x net debt/EBITDA.
TABLE III: PRO FORMA COMBINED SEITEL/PSD
TABLE IV: COMBINED SEITEL/PSD CAPITALIZATION
TABLE V: SVA EQUITY RETURNS
TABLE VI: SEISMIC AND RELATED SECTOR PUBLIC COMPARABLES
Table III presents internal estimates for what a combined SVA/PSD could produce in terms of operating profits. It’s a very simple model but is consistent with recent trends. The outer year projections (2014+) are irrelevant and meaningless and are simply included due to formatting issues. Amortization for Seitel’s data set appeared high based on its latest filing and due to differing accounting practices for amortization recognition between different seismic providers, estimating amortization can be very difficult. Nonetheless, both Seitel and PSD are very similar business models and cash margins for the two businesses have essentially been 75-80%.
The projections in Table III only focus on licensing sales, no revenue from conducting participation surveys is included nor are any assumptions for acquiring other data sets that can contribute revenue factored into the top line. Because of this, sales growth was kept at a flat 3% except for 2008 which was reduced due to deferred revenue recognition for Seitel in the first half of 2007. Seismic data licensing is very cyclical and volatile so it is important to realize that sales can drop considerably over this period. However, the key item to focus on is that seismic data sets produce annualized returns of 30+% over time. Irrespective of short-term macro conditions, the licensing of seismic data should continue to be fairly robust as supply/production tightness can easily result in whipsaw conditions in terms of energy demand. Taxes were left at 0 mainly due to the operating losses due to heavy D&A recognition that reducing operating income and could result in carryforwards that would result in minimal cash taxes required for future years. Since the purpose of this analysis is mainly on cash returns to SVA, significant time was not spent on the tax basis for this transaction.
Table IV illustrates how rapidly the combined entity will delever once PSD’s cash flows are added to Seitel’s. With $75MM in free cash, SVA can utilize that cash to conduct participation surveys which will add to its library and generate future revenues, purchase additional seismic libraries, and pay down debt. Table IV assumes SVA focuses exclusively on deleveraging its balance sheet but PSD shareholders should realize that if SVA elects to conduct surveys or purchase libraries, the overall value of the combined entity would also increase due to the future cash flows from the completed surveys and/or acquired libraries. In addition, the cash balance is maintained as opposed to pay down debt mainly to reflect a balance sheet that had the flexibility to fund surveys or acquisitions immediately as opposed to even utilizing cash flows or drawing on a revolver.
The most important presentation is Table V which illustrates the equity returns ValueAct will enjoy if they acquire PSD for the current $3.10 per share price. Assuming ValueAct closes this acquisition by year-end of calendar 2007 and funds the acquisition with additional debt, the potential returns are extremely impressive in the 2-4 year time horizon. The real beauty of SVA is that, according to the most recent PSD corporate presentation, it commands a valuation multiple of roughly 6.5x EBITDA. If ValueAct can sell the combined entity for the same valuation it paid for Seitel, it can affect a valuation arbitrage due to the cheap price paid for PSD and combining it with its own operations.
PSD investors should keep in mind that the 5.2x EV/EBITDA multiple for the $3.10 bid may also overstate the valuation since PSD recently acquired the 50% of Arcis it did not own which could contribute another $3-$4MM in EBITDA. Additionally, PSD is now in the field conducting a participation survey which would add more data to its existing library, further enhancing the value of the Company. Finally, PSD shareholders would be giving up a real asset in the second largest set of Canadian seismic data for a cheap price that not only produces cash but distributes cash to shareholders in the form of a $0.15 dividend, which has perennially yielded 5-6% against the Company’s undervalued share price in recent years.
Macro conditions and concerns could result in additional volatility for microcap stocks and ValueAct, being phenomenal value investors, may be willing to play off investors’ fears of a recession which could lead to a pullback in oil prices and thus a pullback in the price of PSD shares. Nonetheless, if investors take a longer term, 3-5 year focus, it becomes apparent that PSD could be worth much more than the $3.10 offer price. In fact, running a sensitivity analysis against the previous tables demonstrates that SVA could pay $4.00 per share and still obtain excellent IRRs. In addition, while not a direct comparable, Compagnie Generale de Geophysique (“GGY”) paid 7.0x EBITDA for Veritas DGC (“VTS”) just a year ago.
Table VI lays out publicly traded comps of seismic and related (processing technology, software, etc) companies, further indicating that PSD is undervalued. Earnings are not relevant for seismic data companies due to the massive amortization charges which result in lower GAAP earnings while a high EV/Revenue figure may simply implies that margins for a particular business are high. Aside from EV/EBITDA, PSD is also extremely cheap on a P/TBV value.
While TBV may not be used that often for valuation purposes, PSD’s entire value is based on the massive seismic data libraries it owns. Further, the accrued amortization of those libraries takes 7-10 years while the useful life is easily twice that. Simply stated, the actual economic value of PSD’s seismic libraries are understated from a GAAP perspective which should coincide with a much higher P/TBV valuation, especially when the average annual cash return for PSD’s libraries dating back to 1999 are 32%. In fact, during slower oil periods in 1999-2001, the Company’s annualized returns were still incredibly high, illustrating the quality of the data sets.
At any rate, given ValueAct’s impressive track record, the current offer price of $3.10 ($3.06 when factoring the dividend payment that was announced) is opportunistic and hopefully the current auction run by William Blair in response to SVA’s offer will result in a more competitive bid. Eight parties are currently in the auction process which bodes well for a more competitive offer. In the meantime, shareholders would be better off rejecting SVA’s offer and waiting for the auction to run its course.
Disclosure: Author manages a hedge fund that is long PSD (PSD trades on the Toronto Stock Exchange under symbol PSD as well as an ADR Pink Sheet in the US under symbol PLSDF)