Raven Russia — Update 4 April 2016

Raven’s FY15 results release is realistic about the harsh environment that the company is facing. However, the report also details the company’s active leasing efforts aimed at defending income and cash flow, cost containment, and defensive management of the balance sheet. The year-end cash balance was $202m, a significant proportion of the current market capitalisation, and the decline in underlying pre-tax profit was kept to 13%. This was below our forecast struck after the interim results and we expect the effect to be more pronounced in the current year from the increase in vacancy and market pressure on rents. We show our downwardly revised estimates on page 10.

After reviewing the key features of the 2015 results and the market environment, we also look in more detail at the portfolio structure and leasing developments. Following this, on page 6, we explain in great detail the basis for our investment property NOI estimates, the basis for our group financial forecasts, valuation work, and the key to understanding how sensitive these are to positive or adverse future developments.

As is well known, the Russian economy has been severely hit by a combination of weak oil prices and international sanctions. GDP declined, the Russian rouble collapsed, and inflation increased to double-digit levels. GDP slowed sharply in 2014 and contracted by 3.7% in 2015. However, most forecasters are indicating that much of the adjustment has been made and that the economy will stabilise during 2016 and begin to grow again in 2017, despite the renewal on international sanctions for another year in March 2015. The World Bank (in its January 2016 update) expects a further 0.7% decline this year and growth of 1.3% next; the IMF World Economic Outlook update (also from January) follows a similar pattern of expectations (-1% and then +1%). Inflation has been easing on an annual basis as the worst effects of the rouble devaluation feed through. Annual CPI inflation declined to 8.1% in February 2016 having peaked at 16.9% in March 2015. With the fall in inflation, interest rates too have eased, with government bond rates falling below 10% from a peak of c 17% in late 2014.

The weakness of the economy has the effect of reducing the demand for new lettings and the opportunities to renew existing leases, and has put downward pressure on market rents. The rouble weakness (it has halved in value versus the US dollar since sanctions were imposed in 2014) has put enormous pressure on tenants with US dollar-pegged leases, with some simply unable to cope and others opting for rouble-denominated rents at levels where the US dollar equivalent is well below the levels previously seen in the market, albeit indexed to Russian inflation.

These negative effects of economic weakness were exacerbated by a significant amount of new supply, already in development when the crisis hit, being delivered to market. The majority of this space was developed speculatively with no end tenant in place. In 2014, c 1.6m sqm of new modern warehouse space was completed in the Moscow region (75% of the Raven portfolio) according to JLL. This 16% increase on the 9.6m sqm opening stock had been planned in the context of a structurally undersupplied market with persistent low vacancy and a low ratio of warehouse space per capita relative to other major European locations. New supply continued at a high level into H115 as projects that could not be shelved were completed, adding an additional c 600k sqm, but slowed in H215, taking the annual new supply to 839k sqm. The market vacancy rate increased from 7.9% at the end of 2014 to an estimated 10.5% (on a stock of 12.0m sqm) at the end of 2015. JLL forecasts 700k sqm of new supply in 2016, half of which has either been leased or sold at an early stage of construction, and take-up of 900k sqm. Take-up of space has continued with large retailers in particular seeking to improve their supply chains. E-commerce is expanding with traditional retailers reporting strong growth in online sales, which should support demand for well-located warehouses around the major cities. JLL forecasts a reduction in the vacancy rate to 7.5% by the end of 2016.

Given the developing market situation, from late 2014 warehouse landlords responded by dropping asking prices on new lease agreements in order to protect occupancy and cash flow. Colliers estimates that across the market as a whole asking rents dropped from an average $125 per sqm in Q3 2014 to c $75 per sqm in early 2015. As vacancy continued to increase landlords further responded to the currency-induced pressures on current and prospective tenants by agreeing to rouble-denominated rents at levels not dissimilar in rouble terms to pre-crisis levels, but significantly lower in US dollar terms. Across all sectors rents are for now rouble-denominated for new lettings, although the majority include indexation to Russian CPI.

Rent pressure has been greatest in the market for new tenants, where developers are forced to aggressively compete so as to be able to let space. For existing tenants, contractual lease terms provide some friction to the adjustment process and in lease renegotiations any decision to move rather than renew must include the consideration of the costs and difficulties involved in moving and re-training staff, and possible disruptions to supply chains and existing business.

Raven has a completed portfolio of 1.5m sqm gross lettable area (GLA) of modern Grade A warehouse space and 16k sqm of fully let non-core office space in St Petersburg (Constanta). Contracted occupancy at the end of 2015 was 82% for the overall portfolio and 81% for the core warehouse assets, down from 94% at the end of 2014.The warehouses portfolio is predominantly situated in the Moscow regions (nine assets), with the balance split between St Petersburg and the regions (two completed assets in each). There are no current development projects although Raven does have a land bank of 273 hectares, of which 84 hectares provide land resources for additional phases as demanded at existing completed properties.

Raven’s letting strategy has always been to target what it believes will be the stronger tenants ie, international and larger domestic businesses, which are more likely to fulfil their rental obligations, on secure long-term leases. Diversification of the tenant profile provides additional protection against adverse developments in any particular sector. The largest tenant accounts for 11% of GLA and the top 10 tenants account for 45%.

Over the past year, the tenant mix (by GLA) has shifted slightly towards international tenants (56% of the total at the end of 2015 versus 49% at the end of 2014) with domestic tenants showing a corresponding decline.

In terms of sectors, the most obvious shift in tenant mix during the year was the decline in distribution, falling to 11% (by GLA) from 20%. Tenants in the distribution sector, dealing predominantly in imported goods, were more exposed to the competitive problems created by the rouble decline. Indeed, this is the main issue that all tenants have faced, whether domestic or international, irrespective of sector. The large retailer tenants are an area of relative strength in the portfolio, whether local or the large international players. While third-party logistics appears a relatively high focus for Raven (49%), the majority of this represents large international operators such as DHL, Itelia, DSV, Kuhne and Nagel (as well as Roslogistics).

From the back end of 2014 management turned its focus to renegotiating near-term maturing leases to defend occupancy and cash receipts. As it says in the annual report, management has focused on maintaining income at whatever level, and in whatever currency the market will allow, while protecting cash balances. Exhibit 7 shows the progress that it made on renegotiations and extensions of existing leases, particularly during H1, but continuing through H2. Of the 140k sqm of space maturing in 2015, 80k sqm was renegotiated or extended, as was a material proportion of the 2016 maturities. This exhibit also shows the pick-up in cancellations of leases across the maturity profile, particularly in H2 as tenants with business models that cannot adapt to the new weak rouble/weak growth environment failed or were managed out.

Exhibit 8 provides an update on the opening 2016 position. The 2016 maturities now include the 42k sqm of space that was occupied by Dixy under a PLA at the end of 2015, but which is now vacant. As of 1 January 2016, remaining maturities before the end of 2018 totalled 570k sqm versus 772k sqm as of 1 January 2015.

With the results management reported that since the beginning of 2016 a further 86,800 sqm of renewals and re-lettings had been completed and letters of intent on 29,200 sqm had been signed. The average term of these deals is 3.3 years at an equivalent rent of $73 per sqm, in roubles with Russian inflationary indexation.

Historically, Raven has negotiated rents that are pegged to the US dollar and while the majority of NOI is still US dollar pegged, the proportion of rouble-denominated rents is likely to increase substantially from the 9% share of contracted warehouse NOI seen at the end of 2015 (up from 7% at H1). This traditional model of US dollar-pegged rents effectively transferred FX risk to the tenant but also relieved the tenant of much of the inflation indexation risk, US inflation being consistently lower than Russian. While the growth of rouble-denominated rents will see increasing FX risk revert to Raven, any improvement of the rouble versus the US dollar would represent an immediate benefit while indexation of rents to rouble inflation provides protection against further decline.

The key determinant of Raven’s future profits and cash flow, and ultimately property values and NAV, will be the development of property net operating income (NOI); rental income less direct property operating expenses. There are quite a number of moving parts to this and any future forecast in current market conditions is highly uncertain. However, we have set out in some detail the assumptions behind our near-term estimates (for 2016 and 2017), as well as a potential scenario for medium-term development.

Annualised contracted property NOI at the end of 2015 was $162m, down from $188m at mid-year and $192m at the end of 2014. This is the run-rate of contracted revenues at that point in time and includes the contractual value of any letters of intent (LOI) or pre-letting agreements (PLA) that will not actually contribute to P&L revenues until the lease begins, as well as the internal use of space by Raven’s small logistics subsidiary (Roslogistics), which eliminated from P&L revenues in the group consolidation. We estimate that the difference between average contracted occupancy and P&L revenue was c $17.5m in 2015 versus c $17.2m in 2014. It was lower at H115 (we estimate c $5.6m, largely reflecting the Roslogistic use of space) when the 42k sqm of space occupied by Dixy at Noginsk was included in P&L revenues; given the contractual dispute with the prospective tenant, Dixy, that will result in the space not being occupied this was carried under PLA in H2. The space has subsequently been vacated and we assume it will appear as vacant space in H116, unless it is re-let.

The decline in 2015 NOI, most clearly in H215, reflected a number of different factors:

We think it reasonable to assume that most of those tenants that are unable to meet lease commitments at or around the current $/rouble exchange rate will by now have terminated their agreements. There will be many more who would welcome some relief in their existing rental terms but who are able to meet their commitments within their long-term planning, and for whom the alternative of entering into potentially more volatile rouble-denominated rents and/or incurring the cost and disruption of relocation is similarly unattractive. This leaves the ability to extend maturing leases and/or re-let space, and the terms on which this can be achieved as the most obvious key drivers to NOI from here.

For modelling purposes, we had previously made the assumption that all maturing space would be re-let, at NOI of $100 per sqm (versus an average of c $142 per sqm at the end of 2014), but with a six month lag or void period. To capture the potential impact of tenant defaults we had also assumed bad debts at the rate of 2.5% of NOI pa. The effect of these assumptions was to model a progressive drop in average NOI per sqm while the assumed void period implied a drop in occupancy to 81%. The increase in lease terminations, particularly during H215, as market conditions worsened over an extended period have had the effect of 1) accelerating vacancy; and 2) accelerating the transition towards a lower level of rents.

At this point in time we think it appropriate to adjust our previous modelling framework slightly by:

By way of illustration, we have extended this analysis out to 2020, by which time an implied c 86% of the portfolio would have transitioned to the assumed new rent levels. We call it an illustration because we have not attempted to publish such long-term forecasts and we are conscious that this would be highly speculative, with the actual outcome likely to differ materially, either positively or negatively. Our purpose is to give some indication of the potential direction and speed of travel. We note that we have not increased rents through indexation, even though it is likely to be the case that an increasing proportion of rents will be rouble based and indexed to Russian inflation that is currently c 8%. Implicitly, we are assuming that leases are either US dollar linked (with no growth or indexation) or that the US dollar equivalent of rouble rents will be constant with indexation to inflation equally matched by rouble depreciation. The increase in 2017-20 illustrated average annual NOI is purely the result of assumed improving occupancy as the assumed average level of rents is lower than previously illustrated. We note that with no recovery in rents from the $75 per sqm level to the $100 per sqm level assumed from H217, the illustrated year-end NOI in 2020 would be c $130m rather than c $158m, or c $105m if applied fully to the entire portfolio.

For the P&L account, we have assumed a constant $15m pa deduction from the contracted NOI shown above to make allowance for the Roslogistics internally occupied space and some indeterminate amount of PLA/LOIs that will not immediately generate revenue. However, we no longer deduct the 2.5% of contracted NOI that we previously allowed for bad debts. We note that 2015 was the first year that Raven had ever experienced bad debts and we imagine that these will be at least much smaller or negligible in future given the shake-out of weak tenants that has already occurred.

FY15 results were weaker than we had modelled with NOI c $10m lower than forecast. Underlying cost control was good and associate earnings were strong which limited the EBIT miss to c $6m. However, after finance charges PBT (adjusted for valuation movements) was c $10m below forecast, with a similar shortfall on an underlying basis. Including revaluation movements, the IFRS loss after tax was $192.3m, driving the decline in NAV per share.

We have adjusted our forecasts as shown in Exhibit 11, and we have introduced 2017 estimates for the first time. There are a number of uncertainties and quite a number of moving parts that make forecasting the crucial NOI contribution quite challenging; however, we explain our assumptions and methodology in detail in the section above. Exhibit 12 shows our forecasts for NOI by division.

In 2015, the legacy UK subsidiary Raven Mount successfully sold legacy land plots that it holds in the UK while its JV UK second-home developer enjoyed strong sales. We assume a more modest revenue and profit contribution in future. Roslogistics the third-party logistics subsidiary recorded broadly stable rouble turnover, a good result in market conditions, but saw a 37% decline in US dollar terms. We forecast 10% growth in rouble revenues in 2016 as the company benefits from recent contract wins, but with further negative impact from the average rouble exchange rate.

We anticipate that Raven will benefit from continued cost control although our estimates imply an increase in cost above the H215 level. We have included a notional amount for share based payment charges (excluded from underlying earnings), noting that management is consulting with shareholders on revised management incentive plans that will better suit current market conditions that the old scheme.

We assume no revaluation movements on the property assets and no FX impacts.

The decline in finance charges reflects our anticipation of reducing debt levels, with the cost of borrowing unchanged and constant.

We have allowed for a small income distribution of 1p per year, anticipating that management will wish to maintain a distribution while focusing on debt reduction. As always, we model this as a cash payment although we would anticipate that continuing distributions by way of tender offer are more likely.

We have not assumed revaluation movement in our forecasts, nor have we assumed any FX movements.

The main movement on the balance sheet during 2015 was the $256.5m (c 38 cents per share) negative revaluation movement on property assets, the majority of which relates to the let investment portfolio. Weakness in estimated rental values, the move to rouble rents and rouble weakness versus the US dollar are the main factors driving the revaluation loss. Prime yields are barely changed y-o-y at c 12.0% in Moscow, c 13.3% in St Petersburg, and c 14.5% in the regions. The valuation was undertaken by JLL in accordance with RICS Valuation and Appraisal guidelines.

The revaluation movement is non-cash but it does have the impact of pushing up the ratio of debt to portfolio value, despite ongoing debt amortisation ($57.8m in the year). Exhibit 14 shows that gross bank debt (excluding preference shares) has been increasing versus the (declining in 2014 and 2015) value of (investment and development) property assets. However, in terms of total debt (including preference shares) net of cash, the ratio is actually lower than it was in 2012 – reflecting on the one hand the 2013 preference share part-conversion into equity and on the other, the high level of current cash balances.

New Russian tax legislation that came into force from 1 January 2016 allows the tax authorities to challenge the sale of property assets made by way of offshore SPVs, bringing such sales effectively ‘onshore’ and thereby liable for payment of capital gains tax. Our understanding is that the new legislation is aimed particularly at the use of offshore structures that are in place with the aim of avoiding capital gains tax and the details of how this will operate in practice remain unclear. This is particularly the case with transactions involving genuine offshore company ownership. Nevertheless, there remains a possibility that any such liability could be pursued from the purchaser, and that this could be reflected in underlying transaction values. Given this uncertainty, the underlying NAV calculation no longer adds back deferred tax liabilities in respect of revaluation gains ($49.3m at H115), although there is a hope that the situation may change in future.

Our cash flow analysis in Exhibit 16 shows our forecast that net debt will decline further over the next two years despite the difficult trading environment that we have factored in. The balance sheet forecast in Exhibit 14 indicates that gross debt will fall more than net debt given our expectation that ongoing amortisation at $60m pa will exceed free cash flow.

Our forecasts indicate a stabilisation in ratio of period end contracted NOI to bank interest expense over 2016e and 2017e at c 2.2x (2015: 2.3x).

Raven’s strategy has focused on growing income and shareholder distributions. Currently the strategy is highly defensive; it is still focused on maximising cash flow, but with a greater emphasis on long-term security and debt reduction. It is intended to see Raven through the current crisis and position it to be able to benefit from a market recovery. As a result, we believe that a long-term investment in Raven will remain very much about cash flow and yield.

As discussed above, we forecast ordinary share distributions of 1p per share for 2016 and 2017 after 2p in 2015 and 6p in 2014, before the Russian crisis hit. In part, the forecast decline reflects our expectation of lower profitability and cash flow. But it also reflects prudence as Raven continues to target debt reduction and a healthy cash balance.

Exhibit 17 shows our expectations for operating cash flow net of tax and after interest payments (both bank debt and fixed preference share coupons). While we anticipate a further decline, we expect it to remain healthily positive and well in excess of the modest distribution that we forecast. However, this is before debt amortisation, which runs at c $60m pa, and so any distribution is only possible in our forecasts by reducing cash balances from the current high level, or being able to refinance maturing debt to a higher level. At the current share price this 1p distribution would represent a yield of c 3.3%.

The preference shares earn a cumulative dividend of 12% pa payable on the fixed issue amount per share of 100p. These illiquid shares last traded at a price of 130p, which represents a prospective yield of a little more than 9%.

The NAV per share, on a diluted IFRS basis at FY15 year end was $0.70 (FY14: $0.98) and was the same on a fully diluted adjusted basis (FY14: $1.06). Using a £/$ exchange rate of 1.425 the equivalent fully diluted adjusted NAV in sterling is 49p and the P/NAV 0.60x. However, we do not consider P/NAV to be the main driver of share price or best indicator of value. The investment market in Russia remains relatively immature making portfolio values more volatile.

The duration of the Russian economic downturn is impossible to predict and similarly the future course of the internationally imposed sanctions. We would refer readers to the detailed risk section in the Raven Russia annual report. By way of summary, we would identify the following issues as being of particular relevance.