UK: LBOs – is it terminal or can SSBTs be the answer?

dreamstime_s_94710672

UK: LBOs – is it terminal or can SSBTs be the answer?

UK: LBOs – is it terminal or can SSBTs be the answer?

We have written extensively on the potentially existential risk posed to ‘traditional’ (read all) LBOs by the DCMS Review, especially within the context of profit disruption and the need for a much more developed omnichannel approach than transparent odds, coordinated games releases and (typically poor value) loyalty programmes. Here, we consider the future of LBOs from a product perspective as the results of the DCMS Review now loom (we believe likely shortly after the local elections on 3 May). With all focus on the risk, or otherwise, of shop closures, we consider a possibility more likely and in the long-term potentially more dangerous in our view: a ‘nuanced’ £20 – £2 solution causing a radical diminution in overall profitability creating a ‘zombie sector’ incapable of reforming itself.

Given the focus of the DCMS Review, we should start with gaming, though since we are discussing betting shops we cannot end with it. A ‘typical’ LBO generates c. 57% of revenue from gaming machines (c. £200k), of which c. 60% is from B2 roulette, c. 7% from B2 slots and the remainder (c. 33%) from B3. There was a false working assumption that B3 was going to escape unscathed from the Review; we now know that some level of player tracking (as yet to be defined) is almost certainly going to be introduced (if not in the first wave), given the recommendations of the Gambling Commission. All components of the LBO’s major source of revenue (machines), dropping through to c. 2x EBITDA (post associated direct costs), are therefore highly likely subject to some level of change in the near future (with cross-sector risk contagion likely cold comfort and of limited direct relevance, in our view).

While Gambling Act 2005 is in force and under the current climate, we continue to apply the assumption of four machines per shop; we do not consider the likelihood of a (Local Authority?) right to cut the number of machines per shop more broadly than current compromises (powers over new shops in Scotland and Wales), though this could obviously be worse than even the £2 scenario if introduced (more likely under a Labour than Conservative administration, in our view). Current B3 machines elsewhere than LBOs make c. £3-400 revenue per week, with the B3 element of LBO machines achieving roughly similar levels to the wider industry average (indeed, toward the upper end of the average, albeit negatively skewed in the bingo sector by EBT-machine rollout). If we take a working assumption that substitution cancels out further Social Responsibility interventions, this would suggest an annual machine revenue per shop ‘£2 worst case’ of c. £80k (£400 * 4 * 52 / 1000). A ‘good’ (fairly vanilla £20-£2) review might deliver c. £130k per shop in machine revenue (£625 MWA), with a more restrictive £2-£20 scenario somewhere in the (dangerously broad) middle.

This analysis, if close to the money, has two key direct financial impacts from an LBO operations perspective:

  • The need for high quality B3 content, a functioning supply chain and the cost of SR compliance means that the cost of machines is likely to go up, probably materially (possibly up to 15% revenue from sub 10% currently)
  • A c. £50 – 85k per shop contribution ‘hole’ is created, on average (depending on outcome)
This £50 – 85k figure is dangerous, as all averages are. The ‘average’ shop generates only c. £60k of EBITDA. However, William Hill’s entire estate outperforms this average by c. 35%. Our sector-wide analysis suggests that c. 3,500 GB shops generate more than £80k, while over 2,000 shops generate less than £40k (net of a significant number of recent closures – c. 600 since 2013 with c. 200 added, or an overall reduction of 4% and a ‘poor quality’ reduction of c. 20%). These fourth quartile shops have lower machine revenue in absolute terms, but they are typically more machine dependent, meaning a smaller cash impact is cold comfort in relative terms and risk of closure is far more acute. Further, as much lower revenue shops, their recycling of demand impact on the overall ecosystem is a benefit, but not a significant one, in our view (certainly not a ‘life saver’ outside certain local conditions). Conversely, many successful machine shops have a material high-spending element: these will be hardest hit by any significant utility changes, with lower revenue shops typically (but by no means always) having lower spending customers. Again, this supports the thesis that profitable shops will be hit much harder in absolute terms than lower performing shops – impacting profitability before and deeper than shop closures.

In other words, while a material number of shops are likely to close, the biggest impact of a ‘nuanced’ Review outcome is actually likely to be significantly reducing the profitability of the top three quarters of GB LBOs: they stay open, but they do so on wafer thin margins and become extremely vulnerable to additional shocks (of which there are many: wage inflation, content inflation and channel shift to name just three). In some respects, this is worse than a £2 scenario (still eminently possible, especially if reports of a Conservative back-bench revolt and a Treasury climb-down are to be believed), which will drive more closures, but will at least retrench to a profitable base rather than the far more debilitating medium-term risk (in our view) of ‘zombie estates’.

To put this into context, we see a £2 risk closing c. 2x the number of shops as a ‘complicated’ £20 – £2 and fully 5x a ‘simple’ £20 – £2, but the residual sector profit difference is only 3% between the first two scenarios and c. 75% in the ‘simple’ case (RP estimates); ie, either a ‘simple’ or ‘complex’ £20 – 2 model hits profit far more than closures, whereas the £2 case hits both profitability and closures.

So, what does this really mean for LBOs? That too is extremely simple, in our view: unable to rely on gaming to pay the bills, they will have become principally about betting once more. From a purely retail perspective (and we see any appetite for loss-making omnichannel as likely to be short-lived), this basically requires a combination of two things:

  • Betting product and UX innovation
  • Right-sizing costs
The only meaningful innovation from a betting perspective that we have seen in recent years is the SSBT, dominated by PBS One (BGT – Playtech), with William Hill going it alone. Substantially all shops now have at least one SSBT, with PPB averaging six (and so likely to be similar density in the leading LCL shops). Ladbrokes Coral (now part of GVC) provided data in annual/H1 results which allowed us to piece together its SSBT performance with some accuracy (though it looks like we won’t get FY results due to the GVC merger, meaning H2 results are slightly more guesswork). They show an impressive growth in the installed based and gross win impact, with SSBTs generating c. £90m of gross win in in 2017 from less than £20m in 2014, or c. 75% CAGR to c. 15% of betting gross win Moreover, this has been done during a period in which LCL’s overall betting performance has closed with William Hill. If we consider the period 2014-16 (during LCL’s aggressive SSBT rollout and prior to the TRP disruption, against low levels of WH SSBT rollout), LCL’s OTC revenue per shop (ie, exlcluding SSBTs) fell by an estimated c. £8k vs. WH’s estimated £19k decline. If LCL’s SSBT revenue was materially cannibalising OTC revenue, then we would expect this traditional OTC gap to widen, not narrow (regardless of other management action). Conversely, while WH’s SSBT rollout has achieved a revenue mix of c. 10% in short order, with an estimated run-rate over £300 per shop per week in gross win terms. However, this appears to be much more at the expense of OTC football betting (ie, switching coupon spend), with c. 40% of WH’s football betting now channelled through SSBTs without football betting appearing to grow commensurately (H2 overall betting volumes -3% albeit impacted by very strong margins; OTC volumes -c. 5%). Equally, WH’s performance is still less than half that of LCL, which we estimate at £600+ per week in Q4, or a c. £40m annualised and margin-normalised gross win performance gap on LCL’s estate numbers vs. WH’s performance.

This difference in performance during rollout is important. WH has undoubtedly modernised in-shop football betting UX by moving from paper to digital, but the product range and betting activity appears to be broadly similar, effectively encouraging ‘in shop’ channel shift. This may make retail revenue more sustainable, but it doesn’t necessarily grow it.

Conversely, LCL’s offer is more differentiated to traditional retail (global football multiples from a pan-European provider, broader in-play; especially appealing to younger customers, ethnic minorities and machine players); this seems to have allowed for a far more incremental benefit from SSBTs, coupled with Coral-led improvements to Ladbrokes’s trading policies and a much more aggressive closure programme of an underperforming tail (which WH did not have to the same degree and took action earlier). Overall, betting revenue per shop for LCL appears roughly equal to WH in H217 despite WH’s SSBT rollout: a significant closing of the gap outlined above.

There seems to be two key drivers of this. First, SSBTs, and specifically the relative strengths of the product seem to have led to greater incremental spend in LCL vs. higher levels of cannibalisation in WH. SSBTs do therefore seem to offer some hope regarding retail betting sustainability and growth, but the nature of the product is critical to decide whether this is cannibalising or incremental, in our view. Second, LCL cut its weakest shops, improving the underlying quality of the estate. This is obviously positive, but it also supports our argument that when the DCMS Review impact does come, it is likely to hit the profitability of shops rather than drive substantial closures (again, see above).

Traditional OTC is much trickier. While third-party SSBT content can appear expensive at c. 10% rev share (with outsourced or inhouse servicing on top but broadly similar costs), this is when compared to traditional football (minimal printing and pricing, ex. Sky), which is not necessarily the same product (and strategically should not be, in our view). The mainstay of OTC remains horseracing and dogs: c. 55% of total betting, or c. 60% ex SSBTs, and these have a very different cost base: the horseracing and dog P&L is far more punitive than SSBTs. Factoring in General Betting Duty (15%), the SSBT drop-through to contribution ranges from 65-70% on an outsourced basis (assuming limited horseracing). For racing, factoring in GBD, levy and picture rights, the average drop-through is c. 35%, with the bottom quartile of shops barely breaking-even pre- general overheads (we estimate the minimum horseracing and dog revenue per shop for contribution break-even is £45k). This is not sustainable, in our view.

Shop closures rather brutally resolve this sustainability issue from a purely LBO perspective by distributing similar revenue around a smaller cost base (causing a significant supply-chain issue regarding per-shop-based picture rights, with levy likely far less impacted due to migration). However, the ‘zombie’ scenario outlined above will require creativity, since it effectively preserves supply (short term good for horseracing, though postponing rather than relieving problems, in our view). We believe a combination of three things are likely to happen in this ‘zombie’ outcome: businesses might attempt to run shops without horseracing (contractually difficult but not impossible, especially for ‘new’ operators), with SSBTs being the key attraction (given the limits on machines); racing content contracts will have to be savagely renegotiated (c. 2021-22 without some form of force majeure) and/or retail horseracing and dogs will need to undergo its own retail content and UX renaissance in order to rebalance the relative value extracted to more sustainable levels. This will require cooperation between and within sectors not known for harmonious working patterns; though crises of this magnitude have healed deeper wounds and created stranger bedfellows (again, non- UK – traditional thinking is likely to be vital here).

In terms of other costs, we see little room in the system. Unmanned shops are likely to be regulatory anathema in the current climate, while single staffing also retains its controversy (including now from the RGSB), meaning a combination or potential regulatory pressure and retrenching to better quality shops is likely to put upward pressure on staff costs per shop. Rents are largely set by market forces and while there is downward pressure, this is likely to be gradual. In terms of utilities, the point should be to keep the lights on, lack of management action to tackle that question in the longer-term notwithstanding. We see little in the way of fixed cost mitigation solutions, therefore.

We would also flag the potential risk of HM Treasury being ‘helpful’ in a similar way that it was to bingo in 2009: removing the 25% MGD rate for B2 on a more restricted product could provide the ‘logic’ to move GBD (and therefore almost certainly RGD) to 20%. This is a relatively low-level risk intervention in the short-term, in our view, given Treasury’s apparent understanding of the additional disruption this would cause, but it should not be completely ignored, especially if politically motivated.

Existing products therefore have mixed results, while existing costs are largely fixed and present some highly structural problems (without significant closures, new entrants and/or aggressive renegotiations). While more can be done with SSBTs (delivered well) and something needs to be done with traditional products, the key question is what product (new or better) can keep existing customers and bring in new ones? Answering this requires levels of innovation and consumer testing that does not seem to be present in any material form in any major LBO estates. Without this level of investment and engagement (including genuinely strategic engagement with the supply chain), talk of mitigation or substitution will remain just that, in our view. The risk of a zombie sector with only enough life left to complain about its lot remains a very real one.