Why The Market Is Missing The Point On Ambow Education, A Strong Buy

| About: Ambow Education (AMBO)

This is a follow-up to a past article of mine on Ambow Education (NYSE:AMBO), now that we have the Q1 2012 data in front of us.

The stock fell to $3.21 as of Friday's close, a 31% drop from the July 3 closing price prior to the July 5 earnings release. Clearly, the market thinks that AMBO's performance has materially degraded and the business is neither as robust in its growth nor as profitable as previously thought. The reality is that on an apples-to-apples basis the company had a strong quarter, beating analyst estimates. I will explain in detail why this is true later in this article, after providing proper context.

Since the earnings announcement, three analysts have published reports: BAML, Jefferies, and JPMorgan. The BAML analyst had two consecutive Underperform ratings on the stock but now believes it is currently undervalued by 30% ($4.21 price target). The Jefferies analyst revised down her price target by almost half and now believes the stock is undervalued by 56% ($5 price target). The reason why she downgraded to Hold from Buy was that she was unclear about how to interpret the earnings for this quarter and going forward under the new accounting policies as of the 20F filing: "Revising down estimates given uncertainties in cash basis revenue recognition. ... Upside potential may exist if cash collection and new center ramp-ups faster than expected." She acknowledged that per the old accounting policy of 2011, which PWC -- AMBO's auditor since its IPO -- approved last year, the company would've actually beat revenue expectations: "Based on the old recognition policy, revenue would have been ~RMB437.8, beating management guidance." The JPM analyst revised down his price target slightly to $9 (280% premium to today's price), maintaining a Buy on the stock based on undervaluation and an understanding that the results were largely impacted by accounting changes, not fundamentals: "Around 40% of Career Enhancement sales are through distributors, majority of which will be recognized later this year instead of Q1."

The market was obviously disappointed and confused, assumed the worst and sold. In defense of the sellers, AMBO's management team did a poor job of explaining to investors the financial results relative to the accounting changes and breaking down the operating drivers key to understanding the company's performance. I have spent months studying the company and even I had to read the press release a dozen times to get a sense of what was going on. Let's just say that investor relations is not AMBO's forte, even if operating is.

Having the luxury of being based close to the company, I have spent considerable time doing diligence "on the ground" and will share my research and analysis to clarify several misconceptions about AMBO's latest earnings release.

Let's state the facts, which you can verify for yourself by reading the official press release:

1. AMBO underwent a significant accounting policy change as of May, when it filed its 20-F. Without getting into every little detail, which has been discussed now by all the analysts who cover the stock and my previous article on AMBO, the key takeaway is that the fastest-growing segment of AMBO's business, career enhancement (CE), now recognizes its revenue in a way that front-loads all the operating expenses but delays revenue recognition for over a third of CE segment revenue by at least a quarter. Approximately 40% of all CE segment revenue is derived from contracts with third-party distributors of AMBO's education products for career enhancement (e.g., Java programming courses), and most of these distributors are now classified as "new" (not having at least two full quarters of history with AMBO in cash collection). All that revenue, which was previously recognized upon contract execution, is now recognized after cash collection. As if that was not bad enough, the company now takes a very conservative stance on its bad debt provision, even though historically, 100% of all distributor revenue has been successfully collected in cash throughout the two years that the company has been operating the distributor model. Of course, bad debt provisions are a drag on reported earnings (on a short-term -- i.e., quarterly -- basis), but upon collection, will result in outperformance in the future.

2. AMBO reported Q1 2012 net revenue of $59.9 million vs. analyst estimates of $63.5 million (an 8% miss) and net income of -$10.2 million vs. analyst estimates of $2.6 million. The top line still grew 22% year over year, but missed expectations.

3. AMBO revised its CY 2012 guidance down slightly, from analyst estimates of $337.7 million to a range of $301.7 million to $333.5 million, the midpoint ($317.6 million) being 6% lower than estimates.

4. The IFC, a division of the World Bank, issued $50 million of long-term debt for AMBO's use as expansion capital, to expand infrastructure (opening training centers, mainly in higher-growth Tier 2 cities) more rapidly than the already rapid 20%+ projections.

5. AMBO's CFO, who had been at the company less than three quarters, resigned. He oversaw the 20-F filing, which was delayed and caused the first big drop in the stock, and also oversaw this Q1 2012 earnings release, which was again delayed due to the 20-F delay.

6. A former employee alleged that a small acquisition done in 2008 was suspect. The company's board of directors wanted to fully address any claims of impropriety given the commotion caused by the 20-F delayed filing, and announced an internal investigation to explore the allegation.

Facts stated, let's now review what they mean in light of the realities of what we know about AMBO.

Tackling the easy issues first, I will save the detailed financial commentary for the end, since I believe that is the least understood of all.

AMBO's former CFO did not deliver during his brief tenure. He was brought in to upgrade the company's accounting and audit practices and communications to investors. What he did was the exact opposite. He so mismanaged the audit process that he delayed the filing that caused the huge drop in stock price last month. His commentary during conference calls and various road shows was unclear. I believe that it is highly likely that the board of director's audit committee pressured him to resign and that the resignation had nothing to do with a disagreement with accounting. In any case, I view it as a good thing that Jenny Zhan (currently chief strategy officer and formerly a portfolio manager at American investment firm GMO), who was the effective outward-facing CFO for investors and has done the most to clarify the operations of the company, seems to have taken a more active role in finance to supplement the battlefield promotion of AMBO's accounting director to interim CFO. I view the old CFO's departure as a non-event. He signed off personally on all the company's financials, including the 20-F and the Q1 2012 earnings release before he left and even before PWC signed off, and left with no disagreements as to the veracity or accuracy of the financials.

Regarding the disgruntled employee's allegation, not only is 2008 a long time ago, but the company's audit track record should assure investors that this allegation will be put to bed quickly. The company has had various successful audits, including a pre-IPO audit and two filed 20-Fs since its IPO by PWC, the only American accountancy to never have been implicated with fraud in China. Also, Baring (one of the top P/E firms in Asia), the latest investor in the company, had full access to material non-public information before executing their PIPE transactions and conducted a multimonth audit with a third-party Big 4 accountancy to reassure themselves.

Furthermore, the World Bank/IFC, among the most pedigreed and notable creditors in the world, does not like to issue credit to fraudulent, operationally mismanaged companies that are burning a hole through their balance sheets. Fifty million dollars of credit is a clear sign that the World Bank, after its extensive diligence (typically six to nine months in duration), declared AMBO highly creditworthy. What's more, $20 million of the $50 million was issued as a convertible bond with a conversion price of $10/ADS. Ten dollars, at this point, is a 312% premium from the last traded price. Why would the World Bank want the option to own equity at $10/ADS? Because, after conducting lengthy due diligence with complete access to material non-public information, it believes the business is not only highly creditworthy but also has substantial upside beyond $10.

Earnings Release Analysis

First, if we are going to use analyst estimates as a benchmark for how well or poorly a company performed for a particular quarter, the only way to compare year-over-year or quarter-over-quarter financial results (for any company, not just this one) is on an apples-to-apples basis, that is, with the same accounting standards in place as the benchmark period.

On that basis, here's what an earnings result would look like:

Revenue

  • "Headline" Q1 2012 revenue: $59.5 million
  • Less revenue deferred in 2011 pushed forward to Q1 2012: -$4.3 million
  • Plus Q1 2012 revenue deferred to future quarters: +$14.3 million

Under the same accounting treatment as Q1 2011, Q1 2012 revenue = $69.9 million (a 10.1% beat over the consensus $63.5 million estimate).

Net Income

  • "Headline" non-GAAP net income: -$10.2 million
  • Less NI from deferred 2011 revenue (see note below): -$3.2 million
  • Plus NI from deferred Q1 2012 revenue (see note below): +$10.5 million
  • Plus bad debt provision adjustment: +$2.2 million
  • Plus one-time incremental audit fees (due to 20-F delay): +450,000

Under the same accounting treatment as Q1 2011, Q1 2012 net income = -$0.25 million.

Some notes:

1. The way I calculated net income from deferred revenue is as follows. Because all the operating expenses associated with deferred revenue are recorded real-time (i.e., incurred in Q1 2012), the only two remaining charges to deferred revenue are COGS (which is about 8% average) and tax (about 20%). So in Q1 2012, if you consider that $14.3 million of revenue was deferred but $4.3 million of deferred revenue from prior periods were recognized, there was net $10 million of deferred revenue. Less COGS, that's $9.2 million of gross profit, and less taxes, that's $7.4 million of net income.

2. There was no bad debt provision, and certainly not the 300% increase in bad debt provision of 1Q12 as a result of the new accounting policy, in Q1 2011 (refer to earnings release, which is publicly available). Again, the new accounting policies were implemented specifically to apply a high level of conservatism against the distributor segment of revenues, which were the main focus of the 20-F changes (referenced in my prior article about AMBO). Per the management team during my multiple visits and the channel checks I've done with distributors in Beijing and Shanghai, AMBO generates revenue both directly and through third parties that pay AMBO a fixed fee for the education product and then up-sell, taking profit on the markup. These third party distributors pay after 90 days. AMBO has collected 100% of the time from distributors since using them for the first time in 2010 (this figure from management interviews is corroborated by the minimal bad debt expense historically in all filed audited financials, which is consistent with a similar amount of bad debt in periods before 2010, when the company did not have a distributor channel).

Nevertheless, the new accounting policy asks for all revenue from distributors without a two-quarter track record of cash collections to be recognized upon cash collection and mandates a nearly 4% bad debt provision ($2.2 million on 59.9 million of headline revenue) on total sales (not just sales via distributors). While I am not disputing whether or not the policy is a good one (I personally think it is too conservative given historical facts), the point is that we must consider Q1 2012 in the context of Q1 2011 accounting. Otherwise, the comparison is neither objective nor accurate.

3. The 20-F reveals a sharp increase in the auditor fees for 2011 from 2010, no doubt due to the delayed 20-F filing. The total increase is a little over $900,000, and based on my experience in many such situations (as a client), the bill is split roughly 50/50 before and after the work for lengthy audits that are two quarters long. I've assumed $450,000 of extra expenses in Q1 2012 due to this one-time event.

Training Center Expansion

The most important difference in Q1 2012 vs. Q1 2011 was the opening of 15 new training centers (let's call them "stores"), vs. zero in Q1 2011. That is a material difference to note because a store requires three to four quarters to reach CF breakeven.

Approximately 900,000 RMB per store, or $150,000, of operating expenses are incurred in the first quarter that a new store is added, and very little to no revenue is generated that quarter. So if you want to look at the Q1 2011 and Q1 2012 on a same-store basis, you should net out the $2.25 million of operating expenses that were incurred to launch 15 new stores. That would result in $2 million of net income, slightly below expectations.

But I won't even go there. Let's just say that a 10% top-line beat and breakeven earnings on an investment quarter are very different than the headline results reported by the Street.

New Store Outlook and Seasonality

This industry is highly seasonal. Q1 is by far the lowest-margin quarter because businesses expand most aggressively and advertise the most on Q1, since Q2 and Q3 are peak season in China.

Now let's see what the expansion of stores tells us about the future earnings of the company. Again, we know a store is CF breakeven in three to four quarters (let's assume four quarters to be conservative). Below is the history of new store rollouts since 2010. I've put net margins next to the stores so we can gain some perspective of the cumulative effects of seasonality and startup costs of stores.

New stores added and total net margins for AMBO by quarter:

Q1 2010: 0 new stores added, net margin: 2.3%

Q2 2010: 0, 22.1%

Q2 2010: 0, 10.3%

Q4 2010: 11, 21.6%

Q1 2011: 0, 2.9%

Q2 2011: 14, 21.4%

Q3 2011: 13, 7.9%

Q4 2011: 16, -7.4%

Q1 2012: 15, (apples to apples, not excluding new store costs) -0.4%

Clearly, the first quarter is a seasonally low margin quarter, and the impact of the new stores added dragged margins in Q4 2011 and Q1 2012. that said, the new stores added in Q4 2010 should now be fully operational and generating net cash flow, and by Q2 2011, the 14 stores added in Q2 2011 should too. Furthermore, by the end of the year, the 54 stores added before 2012 and the 15 stores added in Q1 2011 should also be generating regular margins and contributing to cash flow, more than offsetting the 15 stores to be added per quarter in 2012.

When you also consider the reversal of the accounting policies implemented in Q1 2011 in the remainder of 2012 as new distributor sales are collected in cash and new distributors become "old" distributors for whom the new accounting policy no longer applies, I expect 2012 to be another very strong year of growth and earnings for the company.

Disclosure: I am long AMBO.