PPHE Hotel Group — Managing costs well in tough environment

As expected, the COVID-19 pandemic has had a significant impact on PPHE’s results since mid-March, and is likely to do so at least through April and May given the full and partial quarantines that are in operation in the countries in which it operates.

The first quarter of PPHE’s financial year (December year-end) is seasonally less important; Q119 represented 17% of total revenue in FY19. Following a good start to the year in January and February, with total revenue growing by 8.7%, performance slowed significantly from mid-March as quarantines became effective. In March, total revenue fell by 60.2% y-o-y as RevPAR fell by 64.5%, driven by occupancy at 29.6% as the majority of hotels were temporarily closed or capacity was significantly reduced.

Therefore, for Q119, RevPAR fell by 23.3% on a like-for-like basis, which is predominantly due to a reduction in occupancy from 76.4% in Q119 to 58.7% in Q120. On the positive side, the average room rate was broadly stable, in line with management’s trading policy of wishing to hold rates at the expense of occupancy.

There is no quantitative comment from management on trading in Q220 beyond that capacity is significantly reduced in all countries. In the UK, only one of 10 hotels is open: Park Plaza Westminster Bridge to accommodate key workers; in the Netherlands (which is in partial lockdown) four of six hotels are open; and capacity elsewhere is significantly reduced. As March was only partially affected by the travel bans and quarantines, RevPAR is likely to be down by c 90% in April, given quarantines in most countries are likely to be in place for the rest of the month.

The company does not disclose revenue and profitability by country of operation in quarterly trading statements. In FY19, the geographic mix of revenue was: UK 58%, the Netherlands 15%, Germany and Hungary 8%, Croatia 17%, and Other 2%. The geographic mix of EBITDA was: UK 58%; the Netherlands 12%; Germany and Hungary 7%, Croatia 15% and Other 8%.

As is typical in a quarterly trading statement, there is no quantification of profitability for the group. The trading statement on 11 March 2020 indicated that management believed it was not possible to provide guidance and it continues to believe that the outlook is too uncertain to provide any guidance for the year.

Given the slowdown in revenue growth, management has been cutting costs aggressively with a view to maximising cash flow.

In FY19, staff costs of £107m represented 30% of revenue and 46% of all operating costs before depreciation, amortisation and rents and were the company’s largest expense. There is no disclosure with respect to its seasonality or the splits between the different countries, therefore in the absence of this information, the implied simple average monthly staff costs are c £9m. Management is utilising the government support schemes for staff costs, where available, as well as cutting costs by reducing hours, temporary salary reductions, halting contract labour and deferring incentive payments. The chairman of the board and the president and CEO have agreed to a temporary reduction of 100% of fees and salary and there has been a 20% salary reduction for the executive leadership team. As a result, the monthly staff cost is c £5m, with c £3m of this covered by the government support schemes; therefore c £2m is still borne by the company. Consequently, the net monthly cost to the company has reduced by almost 80%.

The trading statement quantifies that the business rates holiday in the UK from 1 April 2020 until 31 March 2021 represents a saving of £1.4m per month, c £13m in FY20, equivalent to 3.5 percentage points of margin on FY19 group revenue of c £358m.

Other cost savings have not been quantified but the statement highlights that a material part of the expense base is variable and, where not variable, that management continues to discuss revised payment terms. We analyse the line items of the operating cost base later.

Cash flow and balance sheet

Management states that the liquidity position is robust given the cash position at 14 April of c £150m and the undrawn overdraft of c £4m. There have been amendments to short-term debt service covenants to ensure compliance with the terms of the covenants for Q1/Q2 and there are likely to be further amendments as discussions with providers continue. At the end of FY19, the company’s cash position was c £153m, therefore cash has been managed well with just a reduction of £3m, and its net debt position was c £515m. The first half of a financial year is typically the least important from an operating cash flow generation perspective: eg H119 represented c 30% of FY19 operating cash flow.

At the end of FY19, the gross bank debt had a term to maturity of 7.1 years, and land and hotel buildings of £888m represented c 56% of total assets.

Management had already announced that a new facility to fund the majority (£180m) of the £200m development cost of the art’otel Hoxton in London has been raised, while also enabling it to unlock £43m of equity that had already been contributed by PPHE. Therefore, of the previously quoted investment pipeline of over £300m, £200m is funded and over £100m has been paused, except for work on planning.

The proposed final dividend for FY19 of 20p per share or c £9m in total has already been cut and in our last update on 19 March 2020 we changed our assumption so that no dividend will be paid for FY20, representing a cash saving of c £16m.

Our new forecasts for FY20 and FY21 are highlighted in Exhibit 2.

On 12 March 2020 we downgraded our EBITDA forecast for FY20 by 14%.

Our key assumption was that occupancy in FY20 would reduce by c 10 percentage points versus FY19 in the key operating countries. This produced RevPAR declines similar to those experienced during FY09, which was affected by the global financial crisis and the outbreak of swine flu. We rationalised this by assuming a shorter but deeper impact on travel in FY20 than FY09. A 10 percentage point reduction in annual occupancy is roughly equivalent to a complete closure of operations for one month or a partial closure for longer, say 50% for two months.

It is now likely that the economic downturn will be greater than that experienced in FY09: the length of the temporary closures, full and partial, of the hotels will be greater than previously expected; and the speed of recovery is more uncertain.

We now assume occupancy levels through FY20 across the group, ex Croatia, will progress as shown in Exhibit 3 to give an annual occupancy level of c 55% and that average room rates will remain as previously forecast. The 85% projection for Q4 compares with occupancy rates in a typical H2 of occupancy in the mid-80’s. For FY21, we continue to assume that all of the lost occupancy in FY20 will return but recognise the uncertainty as to how quickly economies and business and leisure travel will recover. We were assuming a decline in occupancy for FY20 prior to the COVID-19 pandemic given the strong performance in FY19, and therefore are not forecasting a return to the previous peak for occupancy of FY19 in FY21. That said, in the current uncertain economic environment we believe that the risk to our FY21 estimates could be on the downside. We note that an incremental 10 percentage point change in occupancy for an individual month would change the annual occupancy by 0.8 percentage points. A 10 percentage point change in annual occupancy would lead to a 10 percentage point change in revenue, assuming that average room rates are maintained in line with management’s strategy.

For the individual countries, an annual occupancy of 55% would represent a reduction in occupancy of: the UK c 33 percentage points; the Netherlands c 31 percentage points; and Germany and Hungary c 25 percentage points. We will reassess the relative occupancy rates as the year progresses, as the individual countries are likely to perform differently given the different stages of the pandemic and levels of quarantine in them.

We now assume that Croatia will not trade fully through the summer months and will mostly trade successfully for the second half of the summer ie we assume occupancy roughly halves from c 63% in FY19 to 33% in FY20. In FY19, Croatia represented 17% of revenue and 15% of EBITDA, and only operates through the summer months. As a result of this change in assumptions, Croatia represents 13% of the reduced forecast for group EBITDA in FY20.

We hold our prior EBITDA margin assumptions for the individual countries as per our previous forecasts, in which we downgraded margins for the UK, the Netherlands, and Germany and Hungary, but recognise the high level of uncertainty. These imply that 30–40% of the lost revenue flows through to EBITDA for the individual countries. With this assumption, a 10% drop in occupancy leads to a broadly similar drop in EBITDA, if the average room rate is stable, as can be seen in Exhibit 2. Our note of 12 March 2020 highlighted how well the company managed profitability during the FY09 financial crisis, as well as highlighting the flexibility in the cost base. It is clear from the trading statement that management is managing costs aggressively, plus the support offered by various governments for key cost items, such as staff and property-related costs, is greater than during FY09, and should provide support for margins. For FY21, we assume a consistent level of drop through of incremental revenue to EBITDA. Again, we are not assuming a return to peak margins in the key operating countries.

Exhibit 4 demonstrates the individual line items of PPHE’s operating costs before depreciation, amortisation and rents as percentage of costs and as a percentage of sales in FY19. The trading statement has indicated the extent to which internal cost savings and government support have reduced the company’s staff costs, which represent just under half of costs, by c 80%. In addition, the UK business rates holiday is equivalent to three margin points on FY19 revenue. The majority of the other operating costs should vary to a greater or lesser extent with occupancy. Costs such as food and beverage, utilities, laundry and cleaning, marketing and reservation and commissions should have a strong correlation with changes in occupancy. Other costs such as insurance and property taxes, administration costs, maintenance, supplies, IT expenses should be semi-variable relative to occupancy or have an element of discretionary spend.

At present, there is insufficient information to change our assumptions with respect to working capital etc.

To give some indication of PPHE’s seasonality in a typical year, in FY19 H1 represented 43% of total revenue and 37% of EBITDA, therefore it is more dependent on performance in H2, and this is likely to be more the case in FY20.

At 1,160p the shares have rallied strongly from 820p at the time of our downgrade to forecasts on 12 March. The share price performance and our new downgraded forecasts mean that the EV/EBITDA multiples for FY20 and FY21 have increased to 14.4x and 9.0x, respectively. PPHE’s average EV/EBITDA multiple since FY10 has been 8.2x, and in FY09 its peak EV/EBITDA multiple was 19.3x.

The current share price of 1,160p compares with the last-quoted EPRA NAV (end 2019) of 2,546p per share, representing a discount of c 54%.