We generally short companies in four different categories: those who manipulate their earnings through accounting chicanery, those with anticipated liquidity problems (like Tongaat), those with flawed business models (think crypto yield farming), or companies with overvalued stock prices due to a fad (we haven’t seen many of these in SA recently, but we think we will again in the future). The “earnings manipulators” come in two flavours: those relatively good businesses who are simply “massaging the numbers” by using more aggressive accounting policies, judgements and estimates in order to “meet” near term expectations (hiding fundamental weakness in the process), and those “bad actor” businesses who persistently use highly aggressive accounting practices in order to present much higher profits than reality. One of the interesting things for me over the years is how the highly dubious accounting employed by these “bad actor” companies is often a sign that there is something pervasively wrong with the culture of the firm.
The best recent example of this is our successful shorting of EOH Limited, a IT services business that reported 37 straight double-digit EPS growth numbers, which was achieved by massive accounting shenanigans. It ultimately emerged in the Zondo Commission that there was a culture of pervasive corruption in addition to the accounting malfeasance and overstatement of earnings.
What initially piqued our interest in looking at this as a short was the dichotomy of increasing accounting earnings and reducing cashflows. In 2016, while earnings increased by 25%, cashflow from operations declined by 24%, and free cash flow was negative after accounting for capital expenditure. Our analysis of the financial accounts revealed a plethora of issues, from shifting pre-acquisition profits into post-acquisition periods, accounting for 100% of the earnings of acquisitions, but only a quarter of the cost, to things like amortizing internally generated software over an extremely long 15 year period. When we pressed the CFO for an explanation, we were told that “it’s not a cash problem, it’s a cash conversion problem”. The CFO also professed to “not liking” IFRS accounting standards, which is never a healthy sign.
Furthermore, the company’s acquisitions became increasingly bizarre, with one acquisition even providing gardening services, which is quite an odd offering for an IT company. Our own investigations indicated that EOH was keen to purchase almost any business with contracted revenue. At the time of the share price collapse, EOH had 272 legal entities in the group!
The stock started dropping in 2017, first on allegations that the SASSA contract was illegally obtained, by allowing a close associate of a government minister the free use of a R65 million beach house. The Zondo report makes for the most interesting reading on the company’s past, as the new management team did a sterling job of coming clean. Among the highlights was that EOH made a R16 million donation to the ANC’s local government election fund in 2016, a week before the City of Johannesburg awarded them a significant contract. There was further evidence of over R68 million paid in potential bribes related to City of Johannesburg contracts.
The result was that EOH’s run of earnings growth came to an abrupt end, with the 2018 earnings being reported down 70%, while 2019 swung to a loss. In addition to the unwind of the accounting smokescreen, EOH also lost legitimate clients who no longer wanted to be associated with them. We exited with an almost 90% gain on our short. History has shown that we were too early to exit, with EOH stock being down a further 50% in 2023 on a massive R600 million rights issuance, but based on the available information at the time we believed that it was opportune to move onto the next investment idea.