After (mistakenly, in hindsight) selling out of TPW in early 2020, I have been pleased to be able to buy back my shares over the past couple of weeks at an average cost base slightly lower (!) than my previous selling price.
This may sound like anchoring bias (to which I am not immune), but I think the material point here is that the business has grown, over the past two and a half years, into something significantly more valuable than it originally was.
I attach, for ease of reference, a link to the January 2020 post where I outlined my rationale for selling, and why I felt the stock was getting into full-valued territory at the time:
https://hotcopper.com.au/posts/42399363/single
As one can easily see, the ~400m$ revenue target that I considered a reasonable best-case scenario for FY2024 is in all likelihood going to be achieved and exceeded in FY2022; that is two years ahead of time, and without heavily compromising on unit economics.
Of course, like every other online retailer TPW has been a net Covid beneficiary; moreover, after experiencing the initial pull-forward in demand for homewares due to lockdowns and travel restrictions, it has also arguably benefitted from a further boost to sales due to a cheap-debt-fuelled housing boom.
Neither of these temporary effects is likely to repeat itself in the foreseeable future; therefore, also in light of the considerable market share TPW has already achieved, it would seem reasonable to expect a significant slowdown in revenue growth going forward.
Having said that, I also see a number of positives that I think are worth highlighting:
a) Amazon is arguably no longer an existential threat: the US behemoth has remarkably pivoted its business model towards cloud computing services and away from pure e-commerce, with AWS now representing the profitable and growing core of its business.
b) TPW have established themselves as the dominant player in the Australian online furniture market.
c) Online penetration in the Australian furniture market is still lagging considerably behind the corresponding rate in the US/UK; assuming a similar curve to those more mature markets, and by merely maintaining its current market share, TPW would keep experiencing double-digit revenue growth for the foreseeable future (note that this is not necessarily a conservative assumption).
d) According to online traffic analytics website Similarweb, TPW has reached a significant 35% of organic (i.e. unpaid) search engine traffic.
e) According to TPW’s own presentations, the Company is still enjoying a five-year-long uptrend in the conversion rate (i.e. the number of transactions divided by number of unique website visitors in a given period).
f) Because of points d and e (growing organic traffic and growing conversion rate), marketing spend as a percentage of Revenue has remained relatively well-behaved over the past couple of years (currently running at ~13.6% of Revenue), despite a significant increase in Customer Acquisition Costs across the whole online retail sector during the same period.
The key question that is often asked about these high-growth businesses that operate at or close to breakeven by reinvesting all their underlying cash flow into further growth (mainly in the form of marketing spend) is what would happen to the top line if that marketing spend got scaled down to a lower percentage of Revenue in order to try and liberate profit.
Because of points d, e and f above, I would argue that TPW is probably one of the few businesses of its kind that could actually be sustainably profitable, under such a scenario, and at a rate that compares attractively to its current Enterprise Value.
For context, TPW Management estimate that ~75% of total marketing spend is represented by new customer acquisition (mainly via the Google Ads pay-per-click model, as well as any brand marketing initiatives), whereas the remaining ~25% is spent on repeat customers (presumably in the form of targeted ads on social platforms). Within the spend on new customers, the brand-related expenditure (such as TV advertising, etc.) is highly controllable, whereas the cost-per-click side can be adjusted by lowering the bids for relevant keywords when the increase in the conversion rate does not compensate for the acquisition cost.
So, as a first approximation, it doesn’t seem unreasonable to me to assume that a hypothetical reduction in total marketing spend to the tune of -50% (i.e. 25% of current levels used to keep existing customers engaged, and another 25% spent on new customer acquisition) would still see the top line grow in the high single digits. Those percentages can obviously be fine-tuned depending on how the repeat order rate on existing customers compares with the conversion rate on newly acquired customers, etc.
Note that this hypothetical high-single-digit revenue growth target would merely correspond to preserving TPW’s current market share under a conservative upward trajectory for online penetration. Also note that in 1H22 revenue-per-active-customer growth, alone, was already in the order of 10%.
On the other hand, reducing [Marketing Expense]/Revenue from its current ~13.6% to say ~6.8% would liberate, at the current Revenue run rate of ~440m$ pa, some ~30m$ pa of EBITDA; and that, in turn, would take the current EBITDA run rate to ~43m$ pa (assuming a FY2022 EBITDA margin of 3%, as per current company guidance).
Looking at TPW’s balance sheet as of Dec 31st, the Company had ~105m$ of net cash and ~35m$ of net payables (due to their negative working capital model), hence approximately 70m$ of “surplus” capital. At the current Market Cap of 120.5m*3.30$ = 398m$, that corresponds to an EV of 328m$; relative to the hypothetical 43m$ EBITDA achievable by halving the current run rate of [Marketing Expense]/Revenue, that would give an EV/EBITDA of ~7.6x (or a cash-adjusted PE of ~12.0x), for an asset-light business with a dominant market position that would still conceivably be growing its top line in the high single digits.
That looks too low to me, as it is almost an ex-growth valuation, hence why I’ve decided to buy back the stock (regardless of where I had previously sold it).
Needless to say, a reduction by ~50% in [Marketing Expense]/Revenue is not going to happen overnight, if at all; the calculation above should be looked at as a mere indication of whether the business can be reasonably expected to be run profitably at a lower rate of marketing spend, and how that underlying profitability would compare with the current market value.
Looking at possible risks, and beside the obvious looming slowdown in the economy and particularly in anything housing-related, I see it worthwhile asking oneself the following questions:
i) Is TPW’s business model sustainable in an inflationary environment? The Company has so far done a decent job of protecting Gross Margin (still running at ~45.0%, which suggests a modicum of pricing power) and keeping wages under control (~6.0% of Revenue, before stock-based compensation); the main driver of margin erosion so far has been the increase in distribution expenses. The Company’s decision to run a private label business (i.e. with owned inventory), which purportedly commands a higher margin, can be seen as a way of countering those margin pressures; it remains to be seen how it will play out and it is something worth monitoring.
ii) Is the move into home improvement a sensible one? To be frank, trying to compete with the likes of Bunnings starting from scratch did not initially strike me as a great idea. But, if the total addressable market for that is indeed as big as Management claim, and if they can manage to conquer even a small slice of it with decent margins, then it could well be worth it. Fortunately the initial investment into that endeavour is relatively small (10m$), and hopefully Management will have the humility not to pursue it further in the event that it doesn’t prove successful.
To conclude, I am back in with an initial ~1.5%*NAV capital allocation, and will contemplate adding depending on ongoing business performance and price.
As always DYOR, and all the usual caveats apply.
(20min delay)