Foresight Auto — Nasdaq listing ushers in new era

Foresight launched the Alpha version of its daylight ADAS software solution “Eyes-On” in January 2017 (for further details of the technology see our initiation note, Driving the autonomous revolution, 27 September 2016). Even though the company is still working towards its goal of publishing its proof of concept in Q317 (a proof of concept is a body of evidence concerning the product’s performance in a range of conditions), Foresight is already attracting pilot trials from major auto makers in its target markets.

Already being short-listed to compete against big-name players

Foresight’s first trial of its “Eyes-On” system, is with top-10 Chinese auto maker Anhui Jianghuai Auto Company Ltd (JAC). After having selected China as its first international target market and having commenced marketing efforts in Q316, Foresight was invited to take part in a pilot against well-established ADAS players Mobileye and Bosch. We understand that the trial was very successful in terms of the system accurately and quickly identifying potential dangers. If successful, the next step is likely to be JAC bringing FRSX together with its tier one suppliers with a view to integrating FRSX’s software into their ADAS products, with a view to supplying them to JAC.

FRSX believes that if all goes smoothly this could lead to its first order as soon as early 2018. JAC has annual auto production at over 1m units (1.2m capacity), engine production of over 0.5m units (0.8m capacity) and also produces vehicles for European and South Korean manufacturers for the Chinese market. JAC has a leading position in China in the export of commercial vehicles. It exported a total 59,000 vehicles of different types in 2015, principally to Latin American, Middle Eastern and Asian markets. With a leading position in Chinese exports of commercial vehicles we would expect JAC to have an interest in bringing a number of its export vehicles up to Western standards in terms of ADAS related safety features such accident and emergency braking (AEB) and lane departure warnings (LDW). Nevertheless, we believe that a major focus will be on ensuring that its top range products in both China and export markets is equipped with good quality ADAS services, as we believe that better-off Chinese consumers will increasingly expect domestic auto makers to provide similar ADAS features to those already incorporated in mid-range and more expensive vehicles in the US and Europe.

Based on our estimate of 5.5% forward camera penetration for the Chinese auto market in 2018 – still well below our forecast of 18-19% forward camera penetration of US and EU vehicle markets in 2016 – and JAC’s export orientation we see the potential for 2018 orders from JAC to exceed 55,000 units. (See our note Driving the autonomous revolution, 27 September 2016, for more details on our assumptions concerning the adoption of ADAS in China.)

Foresight has also signed similar pilot projects with a second top 10 Chinese vehicle manufacturer, with a similar timeline to the JAC contract, and believes it is close to securing a third, which should further increase the potential order inflow in 2018.

We can see considerable potential for this rapid pace of pilot agreements to lead to an increase in our unit forecasts, based on the assumption of similar order sizes for all three companies. We currently forecast unit sales in China of 9k in 2018 rising to 45k and 162k over the following two years. Nevertheless, at this stage we see it as premature to assume that the company will win the majority of the pilots given the well-established competition from companies such as Mobileye. At the same time it is possible that coordination with tier one suppliers, and delays caused by the need to introduce the ADAS features into new models, etc, may add 12-18 months to the timeline. On this basis, ahead of further concrete indications, we are keeping our existing forecasts unchanged.

With pilot programmes underway in China, FRSX has turned to marketing in the Indian sub-continent. We understand that it is undertaking increasing engagement with auto manufacturers and hopes to gain pilot agreements similar to those in the Chinese market.

The growth of ransomware attacks and high profile infiltration of supposedly secure systems by hackers appear to have reduced public confidence in the concept of fully autonomous vehicles (FAV). A recent survey by the MIT AgeLab and New England Motor Press Association found that the younger demographic, which was more open than older people to fully autonomous vehicles a year ago, is now less trusting of the concept. We see this as important as the technology is still largely optional in vehicles, requiring auto makers to sell the product to a willing public.

At the same time the US Transportation Secretary Elaine Chao has indicated that the US government intends to revise voluntary guidelines set by the Obama administration covering self-driving cars. “Ideas that deliver safer vehicles” have been revealed to be one of the focus points of the changes.

While it is too early to determine how far, if at all, the public mood and government caution may affect the onset of FAVs, we believe that Foresight’s focus on developing markets such as India and China should serve as a significant shield for Foresight over the next two to three years, while the industry hopefully achieves the know-how to secure ADAS networks from attack.

This reflects our view that regardless of the security threat to FAV deployment, auto manufacturers in developing markets are likely to continue to sharply increase “catch-up” spending on features such as automatic emergency braking (AEB), forward collision warning (FCW) and lane departure warnings (LDW) in coming years. Already fitted as standard in most mid- and upper-range vehicles in the US and Europe (see Exhibit 2) and required by EU and US safety authorities in order to score top rankings for vehicles, these features have already established strong track records in preventing accidents and saving lives. Moreover the fact that the driver does not relinquish control of the vehicle with the use of these features makes them less exposed to the concerns surrounding cyber-crime. For these reasons, we expect regulators as well as consumers in emerging markets to continue to put increasing pressure on auto makers to introduce these features, giving a strong incentive to local auto makers to continue with their catch-up investments.

In our report of 24 April 2017, Out of the station and on the self-financing track, we discussed how Rail Vision’s technology has become an increasing driver of value for Foresight. Since then we have increased our revenue expectations for Rail Vision based on firmer pricing expectations for the ADAS units as well as the data that should be derived from the units in operation (rail infrastructure, mapping, weather, agricultural and other intelligence/monitoring). We have also increased our expectations of the company’s potential maintenance income as a percentage of initial sales cost.

We understand that Rail Vision’s discussions with potential purchasers of its ADAS technology support a more confident view of RV’s pricing power than hitherto. In particular, a key potential customer of RV’s systems has indicated that the targeted $70-80k unit price for its ADAS system is in keeping with its pricing tolerances.

We have accordingly revised our forecasts of early unit sales prices from $50k to the lower end of RV’s target price range, at $70k. We continue to forecast a 15% pa decline in these prices between 2018 and 2020 as sales volumes increase to bring the unit price back to $51k in 2019 and $43k in 2020. Thereafter we assume a steady 5% per annum decline as the product matures and competition enters the market.

We also understand that Rail Vision is more optimistic about the potential monetisation of data from telematics equipment once mounted on rolling stock. We set out our concerns in the our initiation note (Driving the autonomous revolution, 27 September 2016) that while there is likely to be significant demand for this data, it is not clear that owners of the rolling stock will not stake a claim for a share in such revenues, reducing the net payment to RV. We have nevertheless increased our forecast of data revenues by from $500 to $800 per train per annum, reflecting what appears to be the potential for higher data revenues than hitherto thought, but also maintaining our assumption of a potential splitting of these revenues with infrastructure owners.

Finally we have re-investigated the issue of maintenance charges, as it relates to our assumption that RV would be able to derive annual maintenance fees representing 2.5% of the cost of equipment sold. Looking into maintenance cost charges for railways on similar equipment and for safety equipment in other fields we now believe that this assumption looks overly conservative. A study by JP Baumgartner into European railway maintenance costs undertaken in 2001 estimated maintenance costs as a percentage of capital cost at 5% for signalling equipment, rising to 10% for telecoms and informatics equipment. Looking at the medical sector, the journal Radiology Business in 2012 observed that third-party maintenance costs for radiology equipment typically averaged 8.3%. Given the common ground in the need to ensure absolute best functioning of the equipment for safety reasons, we have assumed that the group is able to charge 7.0% of its unit sales price in ongoing maintenance charges, putting it at the lower end of the telecoms and informatics as well as radiology fields.

Rail Vision’s marketing efforts have thrown up a wide variety of potential sales prospects in recent months. Deutsche Bahn (DB) remains the most prospective potential client by dint of the level of close work between DB and Rail Vision under the DB incubator programme, in which Rail Vision took part. Israel Railways also has a long cooperation with RV and are thought likely to be one of the earlier companies to place orders. Nevertheless, we understand from management that the company has also been demonstrating products with a major European manufacturer of rolling stock as well as a major European railway with considerable success. The latter already uses remote controls of its locomotives in its switchyards but is looking to RV’s technology to relay risk information to its central control areas to reduce accident levels.

In April 2017, Rail Vision entered into definitive agreements with several private investors pursuant to which these individuals will at some unspecified date purchase an aggregate of approximately $5m of Rail Vision’s ordinary shares at a pre-money company valuation of $20m. The investors will also receive 100% warrant coverage with an exercise price per ordinary share reflecting a pre money company valuation of $25m. The warrants are exercisable until the 18-month anniversary of the date of issuance. As a result, Foresight’s interest in Rail Vision will be approximately 26% and up to 45% on a fully diluted basis as compared with a current 32% and 48%, respectively.

The valuation of the issue at $20m shows the increase in expectations regarding Rail Vision since FRSX’s acquisition of its initial 32% stake in the company in 2016 based on a company valuation of $2.4m.

For the purposes of our DCF valuation, we treat the arrangement as if executed. On this basis we have increased the assumed cash holdings of Rail Vision by $5m to give rise to adjusted net cash pro forma 2016 of $4.2m.

This share issue is slightly dilutive for our valuation of Rail Vision as it was based on a company valuation of $20m as compared with our valuation for Rail Vision of $67.1m (NIS246m) in our last report of 24 April 2017, Out of the station and on the self-financing track. Nevertheless the increase in our underlying valuation since that date due to changes to our forecasts has led to an overall increase in our DCF valuation for FRSX’s stake in RV from $21.5m (NIS78.8m) to $29.1m (NIS106.6m).

Foresight shares commenced trading on Nasdaq on 15 June 2017, using the ticker FRSX. Each ADS represents five ordinary shares. As a consequence of the listing, for the first time investors will be able to read Foresight accounts in English under US GAAP accounting.

Apart from the increase in investor interest that these changes will bring, we believe it is a positive step for the company to commence reporting in dollars. This reflects our expectation that the bulk of product pricing in FRSX’s core business, at least in the next three to five years, will be largely in dollars. We also expect the group to increase its international staffing levels such that the dollar-denominated cost base will also grow.

Quarterly results

Foresight is exempt from the SEC requirement to report accounts on a quarterly basis. We understand that management is likely to use this exemption and will release results only on a half-yearly basis for the foreseeable future. The next results release is therefore expected to be its H117 accounts in August.

The restatement of 2016 earnings from IFRS to US GAAP resulted in a 72% and 86% reduction in reported operating losses and net losses, respectively.

After normalisation of the results losses were 41% higher at the operating and 31% higher at the net level. The most significant factors were:

We have not materially changed our earnings forecasts down to the operating level for the restatement to US GAAP accounts. The key area of change lies in:

As mentioned in our update of 24 April 2017, Out of the station and on the self-financing track, FRSX raised $11.7m (NIS43m) in equity funds in Q117 from the issue of 21.0m shares, and the conversion of 0.1m warrants. Since this point, a further 5.4m warrants have been converted to shares based on the reported increase in capital on the TASE to 100.0m shares as of 27 June 2017. We calculate the funds inflow from the conversions at $5.0m (NIS17.5m), resulting in an estimated $16.5m (NIS60.6m) raised year to date, translating to an estimated $15.7m cash raise net of issue costs.

Management reported on 14 June 2017 that the group has c $17m in cash, which is consistent with our estimate of $14.5m equity raising ytd to that point and a $1.3m cash burn year to date, taking into account the $3.8m cash position at the start of the year.

We forecast FRSX to end the year with $15.8m net cash, based on the above-mentioned $15.7m year to date equity funding, $3.5m operating cash outflows during the year, $0.4m capex outlays and $0.2m inflows from interest income and forex (see Financial summary, Exhibit 11). We expect further warrant conversions to take place over the remainder of the year, if share prices remain above NIS4 per share ($5.68 per ADR), but have not incorporated this in our forecasts.

In our note of 24 April we calculated that the conversion of all the in-the-money warrants would result in a NIS114m cash inflow. Since then the increase in the share price has resulted in all the company’s warrants becoming in the money, including 12.5m warrants with a NIS4 exercise price.

Since that date the conversion of 5.4m warrants has also raised NIS17.5m ($5.1m) for Foresight. We calculate that including a May issue of 4.8m more warrants, current of in-the-money warrants total 53.1m and have the potential to bring in NIS158.2m ($45.0m) on conversion.

Assuming that the share price remains above the NIS3-4 range of the bulk of the options, we believe that warrant conversions should continue to add to the group’s financial strength over the coming two to three years. With our model indicating that Foresight should be self-financing with its current resources, we see warrant conversions as potentially enabling it to exercise its options to increase its stake in RV to 45% (for an outlay of $11.3m) as well as to build a cash cushion for further expansion.

We have increased our DCF valuation for FRSX (see Exhibit 8) from NIS3.30 to NIS3.45 per share ($4.90 per ADR), principally due to the increase in the DCF value of FRSX’s Rail Vision stake, as discussed above, and the inclusion in diluted capital of warrants with a high NIS exercise price (thereby boosting the nominal diluted cash balance). The key fundamental changes are:

Our sensitivity analysis shows that FRSX’s current share price is supported with combinations of terminal growth of 2% and WACC of 9% or terminal growth of a higher 5% with an 11% WACC.

Addressing sensitivity to market shares, our analysis shows the market potentially factoring in scenarios of market shares equivalent to 4% in China and 10% in the rest of the world (ROW) or 25% in China and 5% in ROW. We see the latter scenario as within the bounds of reasonable expectation but it significantly exceeds our current assumptions of 6% in China and 2% ROW.