Background
It was recently announced that the license to run the Irish National Lottery would be put up for sale for a 20 year term. Reports indicated this was expected to raise €400m to €600m for the Irish exchequer. How realistic is that kind of payday for the Irish state?
Finances of the National Lottery
Historical Numbers
To estimate the payment that might be gained, it is useful to consider what sort of revenues and cash flow would be earned by a purchaser. Going back through the National Lottery annual reports provides data from 2002 to 2010.
| average | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | |
| Revenue | €533.0 | €559.0 | €578.0 | €616.0 | €679.0 | €778.5 | €840.1 | €815.1 | €772.0 | |
| Revenue Growth | 4.9% | 4.9% | 3.4% | 6.6% | 10.2% | 14.7% | 7.9% | -3.0% | -5.3% | |
| Prizes | €281.0 | €294.0 | €305.0 | €325.0 | €362.0 | €420.9 | €455.0 | €438.3 | €419.9 | |
| Costs | €79.0 | €78.0 | €82.0 | €88.0 | €99.0 | €112.1 | €117.3 | €113.3 | €108.4 | |
| Cost % | 14.8% | 14.0% | 14.2% | 14.3% | 14.6% | 14.4% | 14.0% | 13.9% | 14.0% | |
| Surplus | €173.0 | €187.0 | €191.0 | €203.0 | €218.0 | €245.5 | €267.8 | €263.5 | €243.7 |
Note that the average revenue growth shown above is just the arithmetic average of the other years shown. The geometric average across these 8 years is almost exactly the same (4.7%). The only set of indicative 2011 numbers I could find quoted show a further drop in revenue, so it seems that the lottery is going to suffer so long as the economy is sliding.
Looking at these high level numbers, it is still not possible to say what the free cash flow to any potential license buyer might be. The chunk of funds available, under current terms, to potentially reward a license holder is the line marked “Costs”, which is currently around 14% of revenue. It is worthwhile looking at how that breaks down, taking 2010 as a sample year.
Of the €108.4m, the cost of goods sold excluding prizes is €86.2m, this covers agents’ commission (48.1m, or 6% of revenue), printing marketing and distribution (19.6m) and “on-line facilities and service costs” at 18.5m.
That leaves €22.2m. The remaining €22.2m is then divided between €19.4m of administration and €2.8m of management fees.
Comparison With UK/Camelot
The UK National Lottery, administered by Camelot, makes a useful comparison with the Irish system given the cultural/geographic proximity, and the fact that Camelot is a likely purchaser of any Irish license.
Camelot returns 40% of funds to the public, based on 28% for good causes, and 12% direct to government coffers. This is higher than the 32% typically returned by the Irish lottery. 50% is paid in prizes (lower than in Ireland, that’s why people say the Irish are lucky I guess). 10% is then kept for administration (this will include non prize COGS). Before getting lost in the numbers, it is maybe illustrative to tabulate this:
| UK | Ireland | |
|---|---|---|
| Prizes | 50% | 54% |
| Good Causes & Govt | 40% | 32% |
| Retailer Commission | 5% | 6% |
| Other Operator costs including profit (retention) | 5% | 8% |
There are some interesting differences here that allow a guess at the shape of an Irish license deal. One key difference not clear from these numbers is that the UK lottery license is not sold for a lump sum, but rather based on a competitive process to minimise retention by the operator while meeting various other qualitative criteria.
Imagining the New Lottery Deal
Let’s look at what shape a potential lottery deal might take. The payout to good-causes is likely to stay at least the same, as has been indicated by Minister Howlin. It would be difficult politically to countenance any other deal where a private operator enters the equation. So that is 32% of revenue accounted for. Changing the prize level is tricky for the same political reasons, even though it is higher than the level seen in the UK. Furthermore, given that sales have been falling, it could be unwise to attract negative publicity in this way.
Leaving the retailer margin unchanged at 6% (difficult politically to change, and generally sticky even in UK), leaves us with the same 8% of revenue for the operator that pertains under the An Post administered deal. From this, the purchaser needs to run their business and take a profit.
Camelot manages the UK lottery with a cost-level of 4.5% of revenue. In the UK this means that they have a profit level of 0.5%. If the same efficiency level could be achieved in Irish operations, that would leave 3.5% of revenue as profit for the operator. At current revenue levels, that would be an EBIT of €27.0m. The EBIT should be a reasonable (probably optimistic) approximation for Free Cash Flow since the Irish Lottery Numbers show higher expenses and have no depreciation allowance, with relatively steady debtors/creditors numbers on the working capital side.
This value of €27.0m is on the low side in relation to valuations being talked about in the media (as we will see below). It may not be easy to drop costs below the 4.5% levels achieved in UK operations since:
- Camelot UK operations have been contestable for some time, and have passed through 3 competitive award processes, and so are probably reasonably efficient.
- The room for synergies between UK and Irish operations is reduced since the Irish lottery will still be run as an “Irish” lottery, with its own TV shows, advertising, and branding
- The existing Irish lottery cost base is rather higher, so 4.5% costs will already demand significant improvements in operations.
Calculating an NPV
We can look at the stream of estimated free cash flows over 20 years and calculate an approximate Net Present Value for the license. To do that, the key piece of data that has not been discussed yet is discount rate. On this point I am unsure what to use. Originally I considered trying to estimate a beta by looking at the correlation between lottery revenue growth and market portfolio, but the data set available for Irish lottery performance is so small that this is unlikely to offer much value.
A rather optimistic cost of capital that has some validity would be the Irish Government 10 year bond yield. At time of writing, that stood at 8.2%. Given that the lottery license is in many ways a license to raise a gambling tax issued by the State, and the fact that Irish Government debt is essentially secured by the tax-levying power of the state and reliability of the State as a counter-party, this has some sense. At least this should establish a floor below the Cost of Capital, while real cost of capital may be higher.
Based on that value, and giving a somewhat generous growth rate of 4.9%, we can calculate the NPV:
| FCF % Rev | FCF Y1 | Growth Rate | WACC | Term | NPV |
|---|---|---|---|---|---|
| 3.5% | 27m | 4.9% | 8.2% | 20 years | 378.8m |
| 4.0% | 30.9m | 4.9% | 8.5% | 20 years | 432.9m |
| 6.5% | 50.2m | 4.9% | 8.5% | 20 years | 703.5m |
Conclusions and Comments
The valuations above make the payday touted in the media look optimistic. Getting to a payout similar to that being discussed (€400-600m) involves making a series of optimistic assumptions:
- Assume financing for the project is available at the same rates as financing for the state at large
- Assume price levels can be decreased without affecting revenue growth (e.g. to get 6.5% of revenue into FCF)
- Assume a restoration of long term revenue growth (to 4.9%) ahead of any prediction of economic growth
- Assume the incoming operator can deliver major cost savings (in fact, the uncertainty of this should drive up the WACC for the project)
- Assume no major investment required, or is covered in the costs used to calculate the EBIT (unrealistic over 20 years, also conflicts with cost-reduction plans)
- Assume taxes are not levied on the corporate profits
These all have a negative effect on the anticipated returns. The only clear possibility left out with positive potential is that there might be growth in the business, but that is probably unlikely to be the case. The current distribution model is well understood and saturated. Innovative online modes of distribution may offer new opportunities, but they also expose the lottery to international competition. In any case, the scope for the lottery is fundamentally Irish, and this constrains the ability to grow.
So, overall the valuations being discussed seem on the optimistic side. Reality may be a bit grimmer, as is usually the case when you buy a lottery ticket.