ArborGen Holdings — Positive actions sustaining EBIT progress

The company changed its name to ArborGen (from Rubicon) at the end of September 2019. H120 results showed good revenue progress and narrower US EBITDA losses in ArborGen’s seasonally quieter US and group first half trading period. Management expects to deliver record revenues in FY20 despite prior year hurricane damage and US GAAP EBITDA (pre central costs) up 20% y-o-y. We have lowered our current-year PBT expectations – dampened at the EPS level by a tax credit – with minor net changes to headline estimates thereafter. The company’s established business development strategy of increasing the volume and proportion of advance seedling sales is very much intact.

While the US is ArborGen’s largest sales territory, its seedling uplift and forestry customer planting season is very heavily skewed to second half trading (October to March). Consequently, the contribution to first half revenues is minimal, though of course local opex (including head office, lease and sales and marketing costs) continues to be incurred. Hence, a US GAAP group EBITDA loss in H1 is the normal trading pattern and the year-on-year reduction seen here in H120 primarily reflects increased revenue and profit contributions from ArborGen’s other trading regions we believe (see later).

During the course of H1, in addition to regular seed orchard and nursery activities, ArborGen will also have continued the development and integration of two nurseries brought into the portfolio in the previous financial year.1 Most obviously, with an annual cycle complete, the company’s own developed seed can be used for new seedling growth for the coming selling season. With this c 45m initial incremental seedling capacity expected to rise to 60m over time as the Taylor nursery is developed, this suggests a level of confidence in the future prospects of the group in this region. The most obvious representation of this comes from the company’s own projections of its expected supply/availability of its more advance MCP seedlings, as shown in Exhibit 2.

While the scale is not shown above, management has stated that it expects MCP seed supply (ie the following year’s seedlings) to ‘more than treble from 90m in 2018 to c 300m by 2025’ as relatively new seed orchards mature and become more productive. Further growth in production is expected beyond 2025 also. The significance of this is that mass controlled pollination (MCP) seedling sales value is typically more than double the highest genetics open pollination (OP) equivalent and therefore a potentially significant driver of revenue growth and profitability.

In the near term, the dip in MCP production (in 2019, with a partial recovery in 2020) is understood to be largely a coastal region effect following hurricane-related disruption to the Bellamy, FL, seed orchard operation in mid-FY19. Naturally, this reduced seed production profile feeds into lower seedling sales in the following year (ie affecting FY20 and FY21). Other MCP seed orchard capacity exists in other states, which is able to serve local market requirements. Given the skew to US sales, the FY20 mix effects will become more apparent in H2; we expect that ArborGen will be able to sustain overall MCP levels seen in the prior year ie approaching 90m seedlings. Bearing this in mind, the company commented at the interim stage: ‘orders are at record levels this year, and we are already sold out in many product categories. Subject to the normal weather caveats, we expect to report record total seedling volumes and revenue for the current fiscal year.’

We interpret this as growth being generated by higher volumes of OP seedling sales. For the record, we have assumed that seedling volumes are in the order of 310m in FY20, of which c 29% are MCP (versus 280m and 32%, respectively, in the prior year). This infers some modest dilution in average selling price in the short term. The strategy of converting OP customers to MCP over time and at the same time increasing overall sales volume is very much intact. Overall levels of demand appear to be healthy currently and delivering record revenues (we believe FY18’s US$46m to be the previous peak) would be testament to this.

Other notable actions seen in H120 include:

Australasia was the dominant group revenue generator in H1 reflecting the annual selling bias and seedling planting cycle of the region. Acknowledging this, the H120 results announcement noted that, ‘with the current year results for Australasia largely in hand, we are focusing on the production and sale of our next year’s crop.’ No further financial information was given.

The NZ Crown Forestry’s One Billion Trees (OBT) Programme2 was announced in November 2018; through grant funding and project partnerships the objective is for non-commercial forestry landowners to plant up to 40% of this quantity of tree seedlings in aggregate over the 10-year period to 2028. (The other volume is consistent with and to come from normal commercial forestry activities.) The initiative began operation in early 2019; the latest data (to end June) indicated that directly funded tree plantings had accounted for just over 15% of the total arising from approvals up to that point. Acknowledging that both approvals and seedling nursery owners require a ramp-up period, ArborGen Australasia – as an operator of seven NZ nurseries (four in the North Island, three in the South Island) – will be benefiting from these increased activity levels. Consequently, annualised revenues have risen as a result of this programme. That said, we understand that the funding mechanism is in the process of changing and the impacts of this, if any, remain to be seen. In addition to seedling sales, ArborGen continues to develop its high-value horticultural services (such as quarantine and tissue culture development) in NZ, with species currently including blueberries, hops, apples, strawberries and raspberries.

In more recent news, the coronavirus emanating from China is having some impact currently on NZ logging and shipment rates. The rate at which economic activity in China returns to more normal levels after taking an extended New Year break into account is indeterminate at this stage. However, given timber rotation periods, we feel that this is unlikely to have a material impact on nursery and planting activity in the near term. In Australia, nursery capacity expanded in 2019, with plans to extend this further in 2020 supported by favourable market conditions and demand projections.

This market typically has a more even planting and selling season with a potentially significant addressable segment of independent forestry owners. ArborGen’s commercial seedling activities are founded upon exclusive licensing and development of the eucalyptus clones of International Paper (since 2013) and more latterly Gerdau (from November 2018) as well as some of its self-cultivated pine varietals (from 2014). The trees of both species are harvested for use in the pulp/paper and packaging industries, with eucalyptus also grown to produce charcoal feedstock used in steel production, while solid pine wood is typically used as a building material also.

This region generates revenues with an annualised run rate of c US$8m pa but is still considered to be very much in a developmental phase for ArborGen, approaching break even this year. In Brazil, ArborGen is largely reliant on contract nursery providers3 for growing its eucalyptus varieties and pine seedlings. In November 2019 it entered a five-year arrangement to lease and operate a nursery at Inimutaba (c 175km north of Belo Horizonte, the state capital of Minas Gerais). This adds c 15m eucalyptus seedling capacity, compared to FY19 sales of c 40m (around 80% of the company’s total seedling sales in Brazil in that year). As in other regions, this now allows ArborGen to have greater control with an integrated supply chain process from genetic development to seedling delivery. The benefits of this arrangement should start to become more apparent in the next financial year, subject of course to increasing penetration of the independent forestry sector. The OECD predicts an improving GDP outlook (with real growth approaching 2% in each of the next two years, from below 1% in 2019) with monetary policy and below target inflation supporting investment and consumption in the country.

At the end of September, ArborGen’s net debt position stood at US$31.6m while IFRS 16 leases on the balance sheet were US$5.6m. The purchase of the previously leased ArborGen Inc US headquarters property towards the end of H1 moved it from a ‘right of use’ to an owned asset at the end of the period and was a material line item in the H120 cash flow statement.

ArborGen’s financial year is heavily skewed towards its second half driven by the selling season in the US, its largest market. This is reflected at the gross profit and US EBITDA levels but smoothed by changes in the fair value of biological assets4 (ie H1 credit, H2 reverse/offset) at the reported NZ EBIT level. Excluding this non-cash EBIT benefit, ArborGen saw a US$2.4m operating cash outflow in H1, similar to but slightly less than that seen in H119. The main ongoing components of this were:

Hence, the overall net outflow was chiefly attributable to non-recurring items including c US$2.6m arising from storm damage remedial costs (including additional seed cultivation and customer compensation payments) and c US$1m deferred retention payments from FY19 exceptionals.

Net bank interest costs of US$1.1m were similar to H119 and, as then, no cash tax payment was required. ArborGen capitalises R&D, categorised as intangible capex spend in our model, amounting to US$1.9m in H120. Lastly, the aforementioned US property purchase represented a US$14.4m cash outlay and a further c US$1m in total was spent on other normal capital items, making total capex of US$17.3m for the period.

After the above items, in the first six months of the year free cash outflow was c US$21m. Excluding the one-off US property purchase the ‘underlying’ movement was c US$7m. This was effectively derived from the seasonal US GAAP EBITDA loss and exceptional P&L cash items (with positive working capital offsetting cash interest and capex).

IFRS 16: Noting that this accounting standard was adopted for the first time in reporting the H120 results, for the record the cash flow impact is relatively minor and largely presentational. Ongoing IFRS 16 lease costs are now split between right of use asset depreciation (although we understand that ArborGen previously treated its US head office lease costs this way, prior to acquisition) and associated interest costs. Capital repayments are now recorded under the financing section and amounted to US$0.6m in H1. Hence, in actual cash terms, IFRS 16 lease outflows – including US$0.1m interest - totalled US$0.7m in the period.

Cash flow outlook: Our revised earnings estimates (see below) feed into a small expected reduction in net debt by the end of FY20 to c US$28m. An H2 increase in revenue and profitability is the normal seasonal pattern, along with an uplift in debtors ahead of collection early in the next financial year. However, we have also assumed that there is an abnormal working capital build in H2 to reflect a recovery in seed inventory as orchard output rises after damage incurred in the prior year. For the same reason, we currently incorporate inventory investment in the following two years, which has the impact of constraining net cash inflows to c US$2m in FY21 and then c US$4m in FY22 with a rising profit profile. We model end FY22 net bank debt to be c US$22m and on this basis, we expect pre IFRS 16 net debt:EBITDA to decline to 1.2x by the end of FY22, from 2.3x in FY20.

ArborGen has long-term US dollar borrowing facilities in place:

Among other things, the third-party facilities above have net worth covenants (variously US$27m and US$24m) and US$2m of the US$7.4m cash on hand at the end of September is restricted and on deposit with Synovus. ArborGen NZ also has a NZ$2m credit facility subject to annual renewal.

Normal US seasonal activity levels, absent further material weather events, supplemented by additional nursery facilities (both annualised effects and operational improvements) are expected to be seen in H2. Elsewhere, an improving Brazilian GDP growth outlook feeding into its primary steel and pulp and paper industries should continue to support progress there, while experiences from the initial phase of the NZ OBT planting programme will inform how it progresses during 2020 and beyond.

Our group revenue projections remain the same though there are a number of moving parts further down the P&L behind some relatively small value movements in our headline profit projections (being larger in percentage change terms in the current year, immaterial thereafter). The key elements of this are as follows:

(Note that lower R&D costs are a benefit to our US GAAP EBITDA estimates but have no impact on reported PBT as they are capitalised.) In addition to the above, our previously published estimates now also include adjustments for:

The overall changes to our estimates are summarised in Exhibit 3. The main impact is seen in the current year and chiefly reflects the underlying central/SG&A changes referred to above, while the EPS movement is slightly dampened by the tax credit seen in H1. Beyond the current year, the net adjustments to the headline reported PBT and EPS lines are not material.

Further out, the seed orchard recovery profile outlined by the company should drive increased advanced seed and seedling availability and sales volumes. We will publish FY23 estimates for the first time following the publication of FY20 results.

In valuation terms the above adjustments have little impact on the operating cash flows of the business and, hence, our previously outlined DCF approach has not changed materially.