We now have the detailed results and have had time to both interact with management and digest the changes. Not revisiting the things I said in the prior article, my conclusion on Rolfes is relatively simple: I still like it and it has just gotten a lot cheaper.
Allow me to elaborate a bit on how I arrive at this:
Firstly, there is a common theme between pretty much all of the accounting errors and restatements: they inflated accounting profits with contra-entries into working capital. I.e. Credit income, debit working capital (either via debtors, inventory or lower creditors).
Why is this important?
Well, that leads me to my second point: Rolfes working capital days have halved. In FY 16, Rolfes’ reported net working capital days were 101 days while the restated working capital days for FY 16 is now actually 51 days (i.e. halved!).
This is not a once-off anomaly, but FY 17 working capital days is now 52 days. All the contra-entries into inflating working capital meant that, although accounting profits looked a bit better, working capital looked a lot worse.
So, while Rolfes may have overstated its profits historically, it was, in fact, understating its cash conversion. Luckily, a discounted free cash flow (DCF) model only looks at cash flows for a valuation.
Hence, while I was adjusting Rolfes’ free cash flows for what I perceived to be high working capital costs, I was in fact auto-correcting it for incorrect accounting profits. Same difference.
Before the restatements, Rolfes had a historical average cash conversion rate of only c.50% to 70%, the Group is now reporting a cash conversion rate of almost 90% (H1:17 was a lofty 86%)!
Once again, why is this important?
Because, once I inputted and adjusted my model to reflect the corrected financial records, the major decrease in Rolfes’ fair value came from two elements that had nothing to do with the restatements: (1) The lower growth being experienced in the Group’s agri- and water segments from the Western Cape drought and Botswana, and (2) My subjective decision to hike my Cost of Equity across the Group to control for negative investor sentiment in the stock. In reality, you can actually ignore (2) as that was a personal decision on my part (risk management at a valuation-level).
Even after doing all of this, my fair value for Rolfes still comes out at c.550cps on a Price Earnings (PE) of c.10.9x (using Normalized HEPS from Continuing Operations). Note that Rolfes’ share is currently trading at c.300cps, so this implies a 12m return of over a 100%! Even if I am half wrong, there is no major downside risk from here.
Is this a fair valuation?
Well, consider a couple of things:
- Omnia (OMN) currently trades on a PE of 16.3x.
- AECI (AFE) currently trades on a PE of 11.9x.
- Neither Omnia nor AECI have such a large exposure to the fast-growing Food Chemicals market nor appear to have such rich IP in the Agricultural and Water spaces. Thus, while controlling for size, Rolfes appear better quality than either of these (in my opinion).
- Rolfes now has a strong, experienced and trusted CEO in place (Richard Buttle, ex-FD of Metrofile — see here for that success story). Likewise, the Group has a strong (and invested) Board backing it (look at the recent director dealings for further indication of support).
- Controlling for the Western Cape drought, discontinuation of the silica mine and the Botswana problems, the rest of Rolfes is actually growing fantastically (especially the Food segment, which grew sales organically by c.15% y/y in FY 17. The latest management change also introduced safer continuity in this segment, which is only positive.)
- I have given nil value to the Colour segment, assumed no upside in the discontinuation in the Silica Mine (likely some assets will be realized at a net profit) and I’ve assumed that the Western Cape drought only breaks in a year and a half’s time. I.e. My model itself is built on conservative assumptions, including a conservative Cost of Equity (adjusted upwards) at 18.9%.
- Finally, the noise may make everyone nervous, but it was accounting noise. Notice that not a single adjustment went through cash flows? Well, a business will always be in business if it has more cash coming in than cash going out, which is exactly where Rolfes is.
Hence, I think that a 10.9x Price Earnings for Rolfes is easily justifiable and my 550cps fair value for the share is not unreasonable.
The curveball trading update, restatements and management change are all part of the course of investing in small caps. No matter what you do, there will always be unknown unknowns that pop-up from time to time.
Firstly, that is why we diversify. No matter how high your conviction, never put it all in one single stock. Always hold a bunch of them. Rolfes proves this point nicely (A Forgotten Safeguard).
Secondly, this is a great stress-test for the Group, its businesses and its Board. While it is nasty to deal with events like this, I actually think that the Group has come out of it stronger. Like I said, we now see that Rolfes is, in fact, a lot more cash generative than we were led to believe. That is a good thing. Rolfes also has better and more rigorous financial controls in place. That is also a good thing. Finally, we have an experienced and trusted CEO in place and, likely, a strong, permanent FD soon. That is also a good thing.
In other words, Rolfes might actually be a better Group after all of this. Yes, I know that is a funny perspective for a stock that has pretty much halved this year, but it makes sense in my mind: stress testing a good business just leads to a better business.
Anyway, in conclusion, I repeat what I said earlier: I still really like Rolfes and, particularly at these levels, I think its share is deeply undervalued. I’ve joined the Board in buying a couple more shares at these levels and, depending on the price action and flows in AWSM Fund, I might pick up some more.