Since last November’s US election sent Donald Trump to the White House and shockwaves through the global bond markets, I have been focussing my efforts into finding highly durable businesses whose revenues are either explicitly linked to inflation, or likely to at least match CPI growth over the long run, with a view to invest in them.
Why? To cut a long story short, it appears increasingly evident to me that, after a decade dominated by monetary intervention, the US is now moving towards more aggressive inflationary policies – fiscal stimulus, deficit spending and the like – to rekindle growth. This new attitude is also finding advocates across other major economies, so much so that topics such as interest rate cuts and quantitative easing seem to have now disappeared altogether from the mainstream media’s talk.
While I do not have a view as to whether said economic policies are going to be successful, what seems highly likely to me is that they will be ultimately conducive to lower real (inflation-adjusted) interest rates. Yes, I did say lower and not higher, but note that I am talking about inflation-adjusted rates; nominal interest rates, for what I know, may well go higher, perhaps even materially.
From an investor’s perspective, low and decreasing real yields can be a nasty beast, because their effect on asset prices is that any inflation-beating (or merely inflation-tracking) income-generating securities get ever more expensive; therefore, I have formed the view that this is (relatively speaking) a good time to look for any residual decent yield in this space.
With this mindset, I have over the past three months accumulated shares of toll road operators, airports, pharmaceuticals, funeral operators, dental practice aggregators, even a plumbing fittings business. And, most recently, I have encountered and analysed G8 Education Ltd.
The childcare sector does look to me like a good example of a highly durable business – i.e. with little or no risk of obsolescence – whose overall revenues are likely to grow at an above-inflation rate for the foreseeable future. I will provide more detail in the sequel as to why I believe this is the case.
Incidentally, this is going to be a very long post, so, for the readers who have followed me to this point and are still interested, please make yourselves comfortable.
Having gone through all GEM’s financial statements since their listing in 2007 (as Early Learning Services Ltd), and having built a financial model for the Company, the first conclusion I have been able to make has been that, up to this point, GEM has been an unequivocally successful roll-up story.
What makes a roll-up story a success?
Aside from exhibiting a healthy rate of revenue and earnings growth, the requirements that I believe must be met, in order for a roll-up strategy to qualify as successful (and not yet saturated), are the following:
a) The Company does need to be enjoying the benefits of scale, i.e. there must be a trend of improvement in EBITDA margins, as a direct consequence of the Cost of Doing Business representing a decreasing percentage of Revenue.
b) The quality of the acquisitions made must not decrease over time, relative to the cost of raising new funding; that is, Management must not be making acquisitions that are decreasingly accretive just for the sake of growing the top line.
c) There must be further room to conquer market share in the sector at attractive prices, and the balance sheet of the Company must be strong enough to provide room for additional growth.
Having defined the [Cost of Doing Business], in the specific case of GEM, as
From looking at these figures, it is evident that the benefits of scale have been achieved over time. The main driver of the decreased CoDB as a percentage of Revenue has been (as it should be the case) the proportional decrease in Employee Expense:
[*] The marginal increase in CoDB/Revenue occurred in FY2016 was caused by an increase in the staff-to-child ratios introduced by the ACECQA (Australian Children’s Education & Care Quality Authority) as of January 1st 2016; looking at a breakdown of the above figures by semester, it is evident that, while there is some seasonality in the Employee Expense, this regulatory increase was fully absorbed during the second half of the year:
[Employee Expense]/Revenue by semester
Jun 2015: 57.52%
Dec 2016: 53.82%
Jun 2016: 58.61% (staff-to-child ratio increase)
Dec 2016: 53.31%
In terms of how EBITDA margins have evolved over time, as a consequence of the above dynamics, the result is as follows:
Finally, to complete the overall growth picture, here are the year-on-year and cumulative revenue and earnings growth rates, since the first net profit was reported in FY2010:
I now want to measure the capital efficiency of the roll-up strategy over time. In order to do this, I am going to look at the differential between ROCE (Return on Capital Employed) and WACC (Weighted Average Cost of Capital), where:
ROCE = EBIT/([Shareholders’ Equity]+[Non-Current Liabilities])
and
WACC = weighted average of [Cost of Equity] and [Cost of Debt].
For the Cost of Equity I will use, for each Financial Year:
[Cost of Equity] = [Next Year Dividend ($ per share)]/[Price of New-Issued Equity ($ per share)]
and for the Cost of Debt I will take the weighted average interest rate paid by the Company during the year, as reported in the financial statements. The advantage of this approach is that it allows one to combine under a single measure the Company’s efficiency in using both equity and debt funding. Starting from FY2010, when GEM returned their first profit, that gives me:
[**] FY2012 WACC was impacted by the Sep 2012 share placement, where 35mA$ were raised at 1.15$/share; relative to the following year’s dividend payments, this implied a spike in the Cost of Equity to an all-time high of 10.43%.
What we can see from this data is that both ROCE (absolute return on capital) and ROCE-WACC (efficiency of capital allocation) are presently at or close to all-time highs; therefore, we can conclude that the quality of acquisitions has not been sacrificed over time for the sake of revenue growth; the use of capital has actually been increasingly efficient.
Also, to double-check the quality of the reported earnings, I found it useful to look at the evolution of the [Net Receipts] to EBITDA conversion over time; I used Operating EBITDA figures (i.e. excluding Other Income) for this purpose:
It is evident from these figures that reported earnings have been strongly backed by cash flow.
Incidentally, I also find it relevant to note that, since the first profit was reported in FY2010, a total amount of 313.08m$ has been returned to shareholders in the form of fully-franked dividends, vis-à-vis total reported earnings of 293.62m$.
When it comes to forming a view as to how far the Company has already grown relative to its potential, it is convenient to make a rough calculation of what its current market share is. As of February 2015 (most recent data point I could find from investor presentations), the Australian childcare market counted 6,585 long day care centres. Assuming this figure hasn’t changed materially, with its current 490 centres in Australia G8 Education represents only 7.44% of the market.
According to its most recent institutional investor presentation (Feb 2017), GEM has a current pipeline of 49 committed child care centre acquisitions that are due for settlement over the next two years; that will take the Company’s market share up to approximately 8.20%.
In absence of a larger aggregator within the Australian childcare sector, i.e. being GEM the leading operator in this space, I think it is fair to conclude that there will be still be room for growth beyond FY2018, balance sheet permitting and everything else being the same. Note that I am not even counting potential further expansion into the Singapore market and/or other international acquisition opportunities.
Let’s then look at the state of GEM’s balance sheet: As of Dec 31st 2016, the Net Interest-Bearing Debt to EBIT ratio was (410.65m$ - 26.47m$)/161.73m$ = 2.375; we then have to factor in the 213m$ share placement to CFCG (Feb 2017), and also the cost and expected return of committed acquisitions due for settlement over the next two years, quantified by the Company as 200m$ at 4*EBIT (see Feb 2017 presentation).
That gives NIBD/EBIT = (410.65m$ - 26.47m$ + 213m$ - 200m$)/(161.73m$ + 50.00m$) = 1.875 for FY2019, consistent with the Company’s intention to keep NIBD/EBITDA within 2.00; this means that, even if 100% of interim Free Cash Flow were paid out as dividends (implicitly assumed in the above calculation), as of FY2019 GEM should still be able to repay all its debt out of cash flow within the following two years.
In terms of operating leverage, GEM’s Fixed Charges Cover Ratio, defined as:
was equal to (161.73m$ + 88.40m$)/(47.07m$ + 88.40m$) = 1.85 as of Dec 31st 2016; projecting the Fixed Charges Cover forward to FY2019, under the assumption of constant Net Financing Cost (as no further funding is needed for the existing committed acquisition pipeline) and constant [Occupancy Expense]/EBIT ratio of 54.66% (as in FY2016), we have:
Therefore, the Fixed Charges Cover is expected to strengthen to an acceptably safe level of around 2x, under the conservative assumption of no additional acquisitions outside the current pipeline and no organic EBIT growth. This looks to me like an eminently sustainable position to be in.
And what about valuation?
At its current price of 4.02$, and counting in the new shares issued to CFCG, GEM’s Market Cap is 4.02$ * (385m + 55m) = 1,769m$; against a reported FY2016 NPAT of 80.26m$, that gives a PE ratio of 22.04, which may not look cheap at a first sight.
If we factor in the 50m$ of additional EBIT that are expected to come by FY2019 from the current acquisition pipeline, assuming constant [Net Financing Cost] and a 30% tax rate we have:
which does not look very demanding to me, given the durability of the underlying business.
In order to get a sense of how this valuation compares to other defensive, highly durable businesses listed on the ASX, I will look at the EV/EBIT ratio, to remove the impact of different capital structure and different taxation regimes (for companies with revenues from outside Australia).
What this means is that, for a hypothetical buyer of the whole business (equity and debt) at today’s market levels, the expected pre-tax cash flow in FY2019 would be over 10% of the purchase price.
Looking at the valuation of other aggregators in highly durable and defensive sectors, this compares for instance with a forward EV/EBIT ratio (projecting operating EBIT forward to FY2019 at the same rate of 2.45%), of 18.20 for IVC (listed funeral operator) and 13.78 for ONT (listed dental practice operator).
These companies too have been successful roll-up stories in their respective sectors, and by taking them as means of comparison I am not suggesting that they are now overvalued. In fact, for the sake of full disclosure, I am a shareholder of both and I still see decent value in them.
Within the childcare sector, the second biggest listed Australian operator is Think Childcare Ltd (TNK), which, according to the same EBIT projections, has a forward EV/EBIT ratio of 13.08, despite having a FY2016 EBITDA margin of just 15.60% (as opposed to GEM’s 22.28%) and a Market Cap of only 102m$. The third biggest listed operator is Mayfield Childcare (MFD), listed on the ASX in November 2016, which has a Market Cap of 35m$ and has given a target EBITDA margin of 19.00% for FY2017 in its IPO Prospectus; MFD reported a loss for FY2016 but, taking the FY2017 EBIT target provided by the Company and projecting it forward to FY2019 at 2.45%, the implied EV/EBIT ratio is 7.39 (it goes without saying that a lower multiple here is a natural reflection of the tiny size and virtually non-existent listed history of the Company).
The biggest non-listed childcare operator in Australia is Goodstart Early Learning, which is a not-for-profit enterprise (the one who took over in 2009 the centres previously operated by the dissolved ABC Learning) and has an estimated market share of 7.2%. The impression I get from comparing GEM with the other ASX-listed childcare operators is that the benefits of scale enjoyed by GEM are also making the Company the “low-cost” operator within its sector, i.e. the one with the highest EBITDA margin.
That would justify, if anything, a valuation at a premium to the sector, and somewhat more in line with a IVC or a ONT; note that IVC has a very similar Market Cap to GEM, whereas ONT is worth only 165m$, which partially explains its lower EV/EBIT multiple.
Assuming for instance a target EV/EBIT of 15 for GEM at the end of FY2019, that would imply a share price of 6.79%, 68.91% higher than today’s closing price. That equates to an annual compound yield of 19.09% over a three-year period; adding to that the current dividend yield of 5.97% gives a total return of over 25% per annum.
Another way of looking at the valuation of GEM is by calculating what sort of Internal Rate of Return it would produce under a Zero Real Growth scenario, i.e. a scenario whereby EBIT increases, after a certain point, exactly at the CPI growth rate.
I have previously stated that I see childcare as an example of a sector where revenues are highly likely to increase at an above-inflation rate for the foreseeable future. This has certainly been the case historically, with an average annual increase in childcare costs by 6.70% in the ten-year period between September 2005 and September 2015
and it also makes sense conceptually; the reason resides precisely in the fact that the majority of childcare revenues is covered by government subsidies. It is politically easy to justify an annual fee increase in line with CPI, irrespective of the actual costs incurred by the Company in running the business, and in the event of corporate costs (mostly wages and rents) growing temporarily at an above-inflation rate, it is also politically easy to justify passing those costs over to the end users (and to the taxpayer).
So, if we assume that revenues (and EBIT) grow at a rate that is greater than or equal to CPI inflation, calculating an IRR under a Zero Real Growth scenario is equivalent to finding a lower bound to what the Company is intrinsically worth on a running yield basis (subject to all applicable risks).
I define the IRR as the discount rate such that the Present Value of all future cash flows (extended out in perpetuity, practically speaking several decades into the future) equates the current market value of the Company.
Under the following assumptions:
- Organic EBIT growth rate equal to the long-term break-even inflation rate (I used 2.45% for this calculation)
- Additional EBIT of 50m$ from acquisition pipeline, by FY2019
- Total EBIT growing at 2.45% per annum from FY2019 onwards
- Constant Net Debt
- 30% tax rate
I see an IRR of 8.85% (493bp above the prevailing long-term Australian government bond yield) at the current share price of 4.02%. Note that the implied FY2019 EPS from this scenario is 29c, which is far lower than the (rather ambitious) 40c FY2019 EPS target recently provided by the Company; therefore, it is definitely a lower bound that is being calculated here.
What remains to be discussed is whether this yield is a sufficient reward in relation to the risks that are inherent to this business.
What are the risks?
1) The main threat that I can see is represented by decreasing occupancy levels; according to the Company’s investor presentations, current average occupancy is between 79% and 80%, after peaking at about 84% in 2013 and having been as low as 75% in 2010. The decrease in occupancy rates over the past three years has been blamed on increased supply in ACT and parts NSW (increasing competition by family day care centres) and general weakness in WA and North QLD (mining sector performance positively correlated with use of private childcare, I guess). While the risk of ever decreasing occupancy rates is real, I would argue that it is also a risk faced by the private childcare sector as a whole; being GEM the highest-margin operator, it is also the best-positioned to survive and look for opportunities in a downturn. Arguably, there could also be a correlation between decreasing occupancy rates and the fact that GEM continues to be able to buy childcare centres at 4 x EBIT or less.
2) The second biggest risk I see is political; because childcare is heavily subsidised by the Government, any change in the structure of subsidies (which is obviously totally out of the Company’s control) would have a major impact on revenues. While an element of randomness is present in this regard, it should not be ignored that possible cuts to childcare spending would make any government extremely unpopular; if anything, if I had to take a guess, I’d bet for higher childcare subsidies in the next few years, rather than the opposite. But, in actual fact, I know nothing except that this is a real and unhedgeable risk.
3) Lastly, there is company-specific risk, i.e. risk that GEM will fail to deliver on their growth execution plan. The spectacular fall of ABC Learning in 2007-2008 is probably still fresh in many investors’ mind, and there could be a perception that GEM, as today’s largest and fastest grower in the same space within Australia, could go down a similar path. Having analysed the financial statements of the Company since its listing in 2007, it has struck me as a business that is prudently managed, with a tight control on costs, and with reported earnings that are solidly backed by cash flow; therefore, I see this residual risk, as well as the previously mentioned ones, as being suitably covered by the yield the Company’s securities are offering.
Based on all of the above, I have been buying shares of GEM over the past couple of weeks, for a total of 3% of my investable capital, and will continue to add on any future weakness, as long as the business case remains intact.
Please note that this is all in my opinion only, written for the sole purpose of confronting ideas and (hopefully) generate some interesting investment debate. Please do your own research, and good luck to all holders.
GEM Price at posting:
$4.03 Sentiment: Buy Disclosure: Held
(20min delay)