Value Investors Club / Universal Entertainment (6425)

Description

Thesis summary:

While not as cheap as it once was (nothing is), Universal Entertainment (6425.T in Tokyo, hereafter 'UE') is still in my view the cheapest gaming stock globally, trading at 5-6x FCF on 2019-20 numbers with significant net cash, for a best-in-class pachinko/pachislot OEM in Japan and a fully-owned integrated casino resort in the Philippines ('Okada Manila'). Previous idiosyncratic risks that held back the valuation - namely, the Wynn lawsuit settlement risk; the large and expensive debt pile; and the reputational risk associated with the former chairman and founder, Kazuo Okada - have now mostly dissipated, removing many obstacles to sell-side coverage and institutional ownership. More importantly, UE is likely to be a big beneficiary of an inflection in the Pachinko/Pachislot replacement cycle in the coming years, post a period of extreme uncertainty that depressed replacement rates to all time lows (for Pachislot

especially). At the same time, the long-awaited Manila casino project is finally nearing the end of Phase 1, meaning reported financials will more closely

approximate the underlying economics of the business, allowing the market to fully appreciate, and value, the Philippines' business properly. If the market doesn't do this near-term, management at UE has suggested they may list the Philippines' business partially on the local market, creating another potential even to unlock

value.

I think fair value is somewhere around 7000 JPY, vs 4330 today, for 61% upside. At my target price, the stock would still be on just 9x 2020E FCF; would be giving no value to 500 unfinished hotel rooms that have been largely completed; no value to the unused excess land in Manila; and no value to the optionality around Phase 2 of the casino project. It would also imply a large discount (on FCF terms) to all the other listed pachinko/pachislot names; and at the very low end of listed Philippines gaming names (despite some structural advantages like owning the underlying land). Simply put, there is a massive margin of safety here on either relative or absolute measures that provides substantial comfort.

The opportunity exists because of the sordid reputation of the pachinko industry (mostly undeserved in my opinion); question marks around legacy issues re Wynn and Okada's acquisition of the Philippines' licence (much more deserved); a lack of sell-side coverage (just one firm); and in general the view that there isn't a hard catalyst now that the Wynn issue has been resolved. To my mind at least a couple of these misperceptions can and will be corrected in the near-term (next 12mos). The stock trades ~$13mm/day, with a $3bn market cap, in Tokyo (note the free float is only ~30%), so it is liquid enough for most all funds.

Pachislot market dynamics:

I will not spend much time discussing the Pachinko/Pachislot market (but am happy to answer any questions in the comments). Suffice to say it is a form of Japanese gambling akin to pinball (for pachinko), while pachislot is much more akin to a traditional slot machine but with a higher skill element (you can easily find good examples of the game on youtube). Gambling is technically illegal in Japan, but pachinko/pachislot have been part of the culture since the 1920s and operate in a legal greyzone that is subject to periodic bouts of regulation, but is highly unlikely to be regulated out of existence (as evidenced by the longevity of the industry and low social cost of the activity).

Due to regulatory change, the last couple of years have been terrible ones for the Pachinko/pachislot market: essentially the government changed the rules in early 2017, making newer machines less ‘volatile’ (to decrease the gambling element in the gameplay), which meant that hall operators purchased less new machines so they could keep more volatile machines in play. This is evidenced by a massive decline in replacement rate, to all-time lows of around 39% in 2017 (I am only showing Pachislot, not Pachinko, below, as it is more relevant to the thesis but the trend is the same for Pachinko as well):

There are a few things to note about the above:

- The average replacement rate for machines over the last 14years is ~80% - this makes sense because most halls require new content to drive customer interest;

- This replacement rate has not varied very much even of late, ie in the 2009-2017 period the average is still 70%. In some years it has averaged above 100%, and many operators are known to cycle new machines into the halls every few months to maintain interest so it is not unusual to see high replacement rates in certain years;

- You will notice the orange line – the total installed Pachislot base – has actually increased over time (ie, Pachislot is taking share from Pachinko). This is very good for UE because essentially they only sell Pachislot (25% market share) versus Pachinko (1% market share);

- Since machines only get licensed to operate for 3yrs (after which they need to be recycled/refurbed/replaced), there is a natural limit to the replacement demand decline – in other words operators will be forced to replace/upgrade expiring machines in the coming years, no matter what. The new regulations – which went into force in 2017 – are mandatory from early 2021, in other words by then the entire fleet will have to be up to date on the new regulations;

- You can see this too in the capex line of the only listed Hall operator (Dynam Japan, listed in HK) – where capex (essentially machine purchases), fell to 4.7bn last year vs 10bn the year before and 12-16bn in the two years prior to that…which ties in with the idea that replacement rate is running >50% below normal run-rates

The key take away is the replacement rate needs to jump aggressively in the next couple of years. Perhaps this doesn’t happen in 2018 but it must begin in 2019 and I think is likely to continue in 2020. Since UE has ~25% market share of the Pachislot market, they will necessarily win when this demand comes back. Looking at historical financials, operating leverage is massive in this business because fixed costs are extremely high:

- The major cost is labour, the vast majority of machines are still largely hand-finished (individual titles, even hit titles, are only ~100k units)

- Content costs are expensed through R&D as incurred (this number hasn’t changed much in recent years) meaning there isn’t incremental content licensing when machine sales go up

- This is why back in good sales years - so, for UE, 2013 and 2016, for example – you saw incremental margins around 65-70% on increased sales

In the below, I assume replacement rate rises to 55% in 2018, and 80% in 2019, and that UE market share contracts to 16% (versus around 25% currently). I make this draconian assumption because the main competitor, Sega Sammy, is likely to win back some share it lost the last 5years due to strategic missteps and remedial investment. But even so, I think the Pachislot business at UE can do ~20bn of FCF next year without too much difficulty (NOTE, I fully tax the Pachislot segment with all the corporate overhead/etc excluding the casino business, to properly reflect the earnings power of the business within the current structure):

Where should this trade? Well, the other OEMs – Sega Sammy, Heiwa, Sankyo – trade at least 7x EV/EBITDA, and near 20x FCF. I think 7x EV/EBITDA would be around 10-11x FCF for UE, for the highest quality Pachislot business in the market (best earnings power, most Pachislot exposure, best track record of share gains), and given the quality of these businesses (essentially cash cows), that seems a very reasonable if not cheap valuation to me. This would imply ~235bn of value for Pachislot alone, on my 2019E EBITDA/FCF numbers.

Okada Manila:

The Philippines casino operation has had $3bn invested in it (at least $500mm over budget) and is finally nearing completion, around 1.5-2 years behind schedule. By year end 2018, ‘Phase 1’ will be fully operational (indeed it is already 70% there), and there will be limited/no more capital requirements for it (note that Phase 1 was entirely financed by the parent UE organization, explaining why it was such a huge drag on reported PnL the last 4yrs). Phase 1 comprises:

- 22ha property, 500 completed rooms (another 500 yet to be completed but the tower for

which has been built), 500 tables, 3000 slots, all fully licensed through 2032 through the

regulator PAGCOR

- 60+ shops, 28 restaurants

- Indoor beach club/night club

Note that the land underpinning the resort (a total of 44ha) is owned, 60% by local Philippines’ businessmen and 40% by Okada Manila (to comply with local regulations), and that this arrangement has been tested in the local courts to ensure the ownership of the 60% is not a dummy corporation.

Simply estimating the earnings power of the business on the scale of the operation today – ie only half of the total site, with no implied value for the remaining 22ha of land, the unfinished 500 rooms, or the potential to double the size of the integrated resort after 2020 – suggests ~45bn JPY of EBITDA in 2019-2020, as per the below:

Where should this trade? Despite the strong growth in Philippines GGR (20% yoy of late, most think it keeps growing 15%+ due to tourism, tighter regulations in China, etc), Philippines casino stocks have been aggressively derated, likely on China worries as well as some idiosyncratic issues (there was a shooting a year ago at one of the competing properties by an upset employee; and land leases have gone up aggressively for some expansion plans leading to fears of margin compression). At least the second part of this is not an issue for Okada Manila, as they own the underlying land, and is thus a competitive advantage. Even so, if we simply pick the low end of comps – ie 6x EV/EBITDA – a level that I think implies around 7x FCF because most of EBITDA converts to FCF – the Philippines operation should be worth ~250bn on 2019/2020 numbers. Note that this accords no value to a) the land bank (22ha on site which presumably will be used to develop Phase 2); b) the optionality around completing the remaining 500 rooms; and c) the potential to double the size of the operation (casino all fully permitted at 2x what is operating today) through development. This would also be a substantial discount to the invested capital in the project (around 330-350bn, by my calculation) today.

Putting it all together:

I get a valuation of ~7k, or 60% above current, as per below:

Why now? There are a few reasons:

- Pachinko/pachislot market recovery: you want to own these businesses when pachislot sales are coming back. As discussed this seems quite likely to happen over the next 2yrs, making this the opportune time to own the business;

- Removal of company-specific issues: there were two key overhangs – the Wynn settlement (ie could UE actually extract the value to which it was entitled?); and Okada risk (he was Chairman until late last year and was being investigated for a number of AML issues). Both of these have now been resolved, and with it, the debt burden removed. This makes the name much more investable I feel;

- Improvement in reported numbers: reported numbers have been burdened for years by thehuge costs at the casino operation (and the overruns, of course). Since this is now mostly complete, the reported financials will improve. Thus, I think the stock is trading not just on 6x FCF next year but also 6.5x reported P/E, meaning it will go from loss-making on a reported basis to being cheaper than any other pachinko or casino name globally (that I can find) when it reports 2019 numbers;

- I think they will increase the dividend significantly next year or otherwise find a way to improve shareholder returns (new management in Japan is receiving stock compensation so is aligned);

- Potential Philippines listing: the company has publically stated they want to list Okada in the Philippines in the next ‘1-2 years’. This isn’t exactly tomorrow, but it is still a considerable catalyst since, as discussed, even if it gets a low valuation in line with comps (despite many more favourable operating characteristics) it would still be hugely value accretive to the UE parent.

At the end of the day the bar is extremely low from here, there is no debt/Wynn/Okada overhang, the pachislot market should go from headwind to tailwind, and at the very least Okada costs will decline substantially (in 2H this year) and perhaps drop out completely in 2019. Yes there is still risk re the Philippines gaming market, and license/AML risk – though none of this is unique to Okada and applies to all the gaming names in the Philippines. On a consolidated basis I have the company at 4x EV/EBITDA next year and 3x the year after – both numbers a huge discount to where either pachinko comps or casino comps trade. My SoTP-derived price target implies 9x 2020E FCF for the combined entity, which is still a lot cheaper than the comps as well – hence my view the stock could easily

double again and look reasonable.

I do not hold a position with the issuer such as employment, directorship, or consultancy.
I and/or others I advise hold a material investment in the issuer's securities.

Catalyst

- Recovery in pachinko/pachislot unit sales over next 1-2yrs

- End of Phase 1 development costs at Okada Manila

- Potential listing of Okada Manila in Philippines

- Increase in divs

- Improvement in reported financials quarter by quarter

- Other stuff

1 show sort by

			2018	2019
Price:	4,330.00	EPS	688	750
Shares Out. (in M):	80	P/E	6.3	5.8
Market Cap (in $M):	3,089	P/FCF	6.2	5.4
Net Debt (in $M):	-600	EBIT	0	0
TEV (in $M):	2,482	TEV/EBIT	0	0