I received an email from a senior guy at a big name hedge fund. They were long Longtop Financial. Longtop is a Chinese financial software company that came public through an IPO (led by Goldman Sachs no less) that has been the subject of a few posts on this blog (see here, here, and here).
I wrote a reply – but never sent it. (I did run through the reply on the telephone).
Longtop has been suspended because they can’t get their accounts out. It is almost certain that the accounts are error ridden – but on Longtop’s minimal release it is impossible to tell how error ridden. This could be a minor problem (say 15 per cent overstated earnings) or a major problem (95 per cent overstated operating earnings, cash balances falsified, cash missing, company with undisclosed debt, stock worthless).
behaviour at stock suspension (taking two days to tell us that they could not get their results out and giving us no other information) suggests but does not prove that the problems will be at the worse end of the range.
As an outline of what may be wrong with Longtop’s accounts (I can’t tell and the company won’t tell us) I reprint the (unsent) reply mentioned above. That letter explains why I thought the stock was a very good near term short.
I have edited the letter to take out some personal details.
Sorry to get back to you so late but I was in court watching the sentencing of Shawn Richard. Shawn is a scammer I exposed some time ago… [Details expunged as they are not relevant to the story…]
I am presuming that the “common interest” as you describe it is Longtop. We are short which means our interests are not exactly “common.”
I am going to discuss Longtop and I hope these discussions do not get back to the company (at least with my name on them). However even then I will be circumspect. The problem is morphology of sin. Clean accounts, like virginity, refer to one state only. False accounts, as does sin, refers to many states. Virginity is easy to describe and relatively easy to determine with certainty. Sin – well – that is complex to describe and comes in many variants. Sin has morphology.
With Longtop there is something clearly wrong with the accounts – but it is awful hard to determine what.
Here are the things we know:
1). It generates cash like there is no tomorrow and yet it went to the market and raised cash. It currently has cash equal to 250 quarters of capital expenditure and 26 quarters of all expenditure. By contrast Microsoft – a company with way too much cash burning a hole in its pocket (witness Skype) has cash equal to about 13 months of all expenditure. The company is – on the accounts – absurdly cash rich.
2). The last time I saw a set of accounts like it was CCME which had cash on hand of $170 million and quarterly capital expenditure of about 200 thousand. Obviously the cash was fake there.
3). The company claims a margin at the very top of software companies globally which indicates that it does something proprietary. Proprietary stuff requires protection – you would expect to pay the staff well, have retention processes to keep them from going anywhere, have racks of servers and different computer systems to test/develop your software on protect your product from being stolen. You would have some capital expenditure per incremental staff member.
4). The company claims almost no incremental capital expenditure per staff member. The last nine months are showing capital equipment equal to about a laptop per individual staff member (if that). Previous years have shown incremental capex per incremental staff member of 1500 dollars or so. This is too low a level of capital expenditure per staff member to be doing genuinely proprietary stuff. (Just think how you would be running a business growing that fast doing proprietary stuff based on the intellectual property of your staff. Actually you probably do run a business that looks like that… and the numbers are quite different from Longtop.)
5). It’s possible to have very low incremental capital per staff member if you work like the computer outsource houses of the pre-internet era (the so called “body shops”). In the 1980s a staff member would turn up for a few days and hot-desk at the office whilst getting ready for their next assignment (say to a bank to work on their computer system). Then they would go out for a year working on the bank’s computers and with the bank’s air-conditioning, desks, carpet, vehicles etc. This requires almost no incremental capex per staff member. However it is a 10 per cent margin business.
6). Conclusion: either the capex number is wrong or the margin is wrong (or both). I do not know which though. Its the morphology of sin problem.
7). The Chairman’s behaviour with his own stock is very strange. Bill Gates (someone with what on the accounts is a comparable business) gives his stock to charity. This guy gives it to his staff – and otherwise gives it away. There is a possibility (though it is a wild guess) that the cash on balance sheet which was necessary for the audit came from sale of stock but won’t be there at the end. I honestly do not know. I am just looking at the very strange business of the Chairman giving away his stock. Again I refer to the morphology of sin problem. I figure – as per 6 above – that the accounts are wrong – I just can’t tell how they are wrong.
8). The company has done lots of acquisitions but not diluted its ratios. That is very strange. The ratios are at the very top end of the world of software companies – really extreme. Its pretty hard to mix anything with margins like that and keep margins like that.
9). There are serious difficulties tracing former staff members using the usual channels like Linked In. When you find someone on Linked In who lists Longtop in their resume their new job seems too low in status given their title at Longtop. Alternatively you find people who give very senior positions at Longtop (China General Manager) who have almost no experience beyond a mid-ranking graduate at a Canadian technology school plus one year work experience. This seems odd.
10). There are some known and disclosed low margin businesses. For instance. the last 20F filing contained the following.
With our system integration services, we assist clients with the procurement and installation of hardware and software which best meets their system requirements. We assist our clients in managing the equipment manufacturers, obtaining bids and proposals on their behalf, negotiating terms and where required monitoring the installation and testing, which is normally provided by the manufacturers. In some contracts, we may also provide financing to our clients. Where warranty is required, we obtain, on behalf of our clients, manufacturers’ warranties and support for the third party hardware and software. On behalf of our clients we procure equipment from major international technology companies, including BEA, BMC, Cisco, Dell, Diebold, EMC, Hewlett Packard, IBM, Microsoft, Nortel and Oracle.
Our warranties include service for both hardware and our and third-party software solutions. Although we arrange back-to-back warranties with hardware and software vendors, we have the contractual responsibility to maintain the installed hardware and software. Most of our contracts do not have disclaimers or limitations on liability for special, consequential and incidental damages nor do we cap the amounts recoverable for damages.
This is almost by definition a thin margin business – but the aggregate margin is breathtakingly fat. It is almost impossible to find businesses which are fat enough margin to mix with businesses like the ones described in the quote above to get a margin as fat as Longtop.
Also some of the acquisitions look like “body shops” (as described above). Those are thin margin businesses and again when mixed with Longtop the aggregate winds up as a fat margin business.
11). The banks themselves in their accounts have been showing lower capital equipment and outsourcing spend. Other suppliers in China are complaining. Longtop is not. Its a pretty thin channel check – but it indicative of something being a little unusual. (Again it is hard to find out what – but it goes to morphology of sin).
If I had to guess I would think that revenue is faked (see 11), margin is faked (see 10), hence cash on the balance sheet if it is there (see 7) does not have the source the company would want you to think it has.
The auditor is Deloitte. Audit is due mid June. Delloite have got sudden-wariness since the CCME debacle where they signed off (2009 year) on falsified cash balances. My guess is that the company does not pass audit and is a very good short indeed with a 34 day lifespan from here.
But that is a guess. I am faced with a morphology of sin problem. The accounts are clearly not right (see 6 above) but I am not sure in what way they are not right or the extent to which they are not right. Unlike CCME there is a real business here clearly worth something. When it fails its audit (as I think but cannot be sure it will) then there is residual value (if you can get your fingers on it). If you can get rid of the very senior management (who have been selling the Chairman’s shares fast and who are probably responsible for this mess) then you have a business which has value. This is not a “zero” but it is awful messy. Its a zero if the management have also stolen assets like the hundred plus million raised in the secondary.
If the accounts are only modestly incorrect you have only modest problems. Again its about morphology of sin. Your 17 year old daughter may not be a virgin any more – but that does not mean she is going to hell in a hand-basket.