Spectra Systems - Developing Aeris

http://spectra-science.com/


I first covered US-based Spectra Systems in March when I noted that "despite the lacklustre share price performance, and the illiquidity of the stock, the market looks to be severely undervaluing Spectra's prospects". Nearly six months have passed now, and through the release of several RNS' there is now a new set of information to work with, that helps to update the investment case for the company. The share price has since advanced from the circa 18p-19p range when an overhang was present, to 28p currently, yet that is still shy of the 75.3p (£34m market cap) upon listing in 2011 when Spectra successfully raised net cash of almost £13m. Despite cash being deployed during the period, there are reasons to be highly optimistic that the share price rise has not run out of steam.

This article is a follow-on from the below post. For full context and information, read this article first:
http://www.theel1tetrader.com/2014/03/spectra-systems-cleaning-up.html

Technicals


The technical outlook for Spectra has strengthened significantly since the original review. At the time, the ask price was close to the 18p mid-price by virtue of a stock overhang from an institution wanting to trim their holding. That presented an opportunity to essentially purchase stock at an attractive discount to the normal spread. That overhang eventually shifted after a few trading sessions, and the share price rapidly rose towards the low 20p range. The progress since then has been highly positive and a reflection of the good news the company has released - the relative ease at which the share price has risen is particularly pleasing as it supports the view that there is not an overhang and that the large number of institutions holding the stock do not wish to sell into the rise, even with the slightly better liquidity. The spread is currently 27p-29p, which I maintain is not unattractive given the current valuation.

There have been several institutional movements since the previous review. Artemis Investment Management reduced their stake to under 5% of the shares in issue, whilst the City Financial UK Equity upped their stake to over 6% and Spreadex increased their percentage stake (through a CFD/Spreadbet) to 4.6%.

Operations

In the initial review I covered the three core aspects to Spectra's operations; banknote materials and products, internal control systems and Aeris, the 'revolutionary' new product that Spectra had developed to clean banknotes through the use of supercritical CO2. Further progress has been made on all fronts since then, and they are all largely encouraging and help to bolster the bullish case for Spectra's future.

First off I'll re-cover the main points that make Spectra attractive when discounting Aeris. In the banknote division, the company is involved with the production and selling of vital materials for both banknotes (e.g. phosphor and security components) to central banks worldwide. Therefore, the type of customer that Spectra sells to is obviously very high calibre, and the earnings quality on the repeat sales for that is high given the multi-year and in some cases, multi-decade contracts. Products sold include banknote authentication systems capable of reading tens of banknotes per second, scanning them to isolate counterfeit notes. The central banks involved include two within the G8, 17 others for currency authentication and a major G8 bank for passport security tools. That is the existing ongoing business that has been stepping up in scale over the past few years as a result of a multi-million dollar sensor contract and organic growth.

The Internal Control Systems (ICS) division adds a more recurring side to the company through the provision of real-time fraud control and risk management systems to numerous customers in 9 countries. Within that figure, ICS provides systems to multiple US state lotteries including Idaho and Florida, and most recently in April when the company executed a 4 year contract with extensions for Michigan's state lottery - being based in the US is certainly advantageous in this respect.

However, it is the Brand Authentication (BA) division which is worth a closer look. As commented upon in the initial review, Spectra completed the acquisition of Inksure technologies back in March of this year having begun the progress in H2 2013. Although the headline acquisition cost of circa $1.36m may at first appear steep since Inksure was not profitable and burned close to $1m during the last financial period, there were plenty of synergies and the removal of resource duplication actually means that the new BA division should contribute materially to revenues and profits over the coming years. Moreover, brand authentication is a potentially interesting growth sector for Spectra - whilst the company notes that there have been several failed attempts to properly exploit the opportunity in recent years, there was an encouraging initial sign in the form of an executed $100k+ contract signed in July. That is just months after the completion of the acquisition.

Work has since been done to develop and move towards launching a new brand authentication application that will provide direct exposure to consumers when released. Named TruBrand, the application enables consumers to detect counterfeit goods through scanning the security features shown on those products. For example, assume that you wish to purchase an expensive bag and see it at a discounted price, but you are not sure it is legitimate. Assuming it is the right type, you photograph the security feature using either an Apple or Android device. In one case, the photo will be used to identify whether taggants exist (optical codes not visible to the naked eye) - if they do exist within the foil, then the smartphone application will let you know that the product is genuine. These security features are sometimes implemented into the product directly too, meaning that counterfeit goods are much easier to identify, just through using a smartphone application. If signed up for, the smartphone can also send data to the cloud where authentication is determined.

Whilst the $100,000 in revenue is not particularly material, this is the contracted revenue and there is scope for "significant growth potential." More importantly, it is an early validation of the technology, and although it is obviously not conclusive, it bodes well. It's key to remember that Spectra are at the forefront of innovation within their sector, and that is largely down to key staff members who have significant expertise. Aside from the direct use of the tool, it actually also allows for incremental revenues to be possible through offering cloud data access and data mining services (e.g. information on how many counterfeit goods for a particular brand are found and if there is a concentrated location - that is valuable information for the brand owner). The contracts are with two suppliers in Asia - the Korean contract is exclusive whereas the Chinese contract is non-exclusive, but both include recurring licence fees and minimum annual order quantities. The contract also makes financial sense, and although the minimums are not huge, the net margin on recurring revenue from the contracts is likely to be approximately 90%, so accounting for that, it is far more material. Indeed, it is equivalent for a contract size with revenues multiple of $100,000, but with sub-50% margins.

Before looking at Aeris again, the heart of the future potential, it's worth noting again the new 25,000sq ft manufacturing facility that has received quality control approval. Multi-year contracts are expected to move over to that facility after final vetting from central bank customers is complete, and that should enable margins to grow by as much as 15% on corresponding sales. Crucially, the following contracts are designed to absorb a major proportion of the facility costs, therefore allowing Spectra to recoup some of that expenditure.

However, it is Aeris that is the most interesting aspect of the company at the moment - the aspect that Spectra touts could drive over $1 billion of hardware sales with ongoing service revenue. It doesn't need to be said that Aeris carries huge potential. I won't cover all the parts of Aeris that I did in the initial review, but rather cherry-pick the most interesting details and add further information.

It is the scale of the opportunity for banknote cleaning that is highly attractive - Spectra has essentially created its own market. "Aeris has no competition, and once patents are granted, will be our market exclusively." The Aeris patent is currently going through a fast track program through the US patent office and is expected to be issued in Q1 2015 - that could prove to be a major catalyst for the share price. Spectra is also in discussions with international banknote equipment manufacturers to handle the Aeris service functions on a worldwide basis.

The Aeris system carries the potential to generate billions of dollars of global savings on a per year basis according to the company. Through cleaning the banknotes, money is saved through being able to effectively 'rescue' those that were too soiled to use. Money is therefore saved on printing costs. Money is also saved through not having to shred as many old banknotes, and further on not having to pay for landfill or other disposal methods, which are not environmentally friendly. So there is a two pronged advantage for central banks - reduced costs and reduced environmental impacts. To a lesser importance, the cleaning method can remove bacteria and motor oils and extracted oils and sebum are 'ideal additives' for diesel and heating oils. Two of the most advantageous points about Aeris is that the notes can be cleaned even whilst strapped, shrink-wrapped and bundled, and the cleaning process does not damage security features, nor alter the structural integrity of the note.

Credible progress has been made over the past six months, over which the number of testing programs entered into has doubled from the two previous G8 Central Banks, to also include Banco de Mexico and the Central Bank of Indonesia. Further testing programs are expected to be announced later in 2014, so there is plenty of scope for Spectra to showcase their product using the current prototype in the testing programs. That comes before Spectra announced a partnership with Cool Clear Technologies in July for the construction of a fully productised unit at a development cost of $400,000 for delivery in November. Two unit types will be explored for 300,000,000 notes per year and 600,000,000 per year. The designs will be suited for a 10 year life cycle that requires routine preventative maintenance and the company has already noted that various tests with central banks have generated "very positive results".

Commenting on the testing programme with Banco de Mexico, CEO Nabil Lawandy commented, "We are very pleased that the Banco de Mexico has provided us with the opportunity to collaboratively assess the efficacy and economic impact of our banknote cleaning technology for Mexico's banknotes. Mexico is one of the countries which has adopted polymer notes for its lower denominations and has had extensive experience with the issues of soiling and reducing banknote replacement costs. In addition, the Banco de Mexico has one of the most highly respected central bank research and development teams in the world and has developed important data on banknote life cycles. We are extremely excited about the opportunity to gain their insight on the value of our breakthrough technology."

The sole access to the market will enable Spectra to achieve very attractive margins either through outright sale of the systems or through leasing them out to central banks and associated companies.

It's when you start looking at the numbers that the opportunity becomes clearer. In the USA alone, typical note production volume is 4-6 billion per year and the higher denomination notes are more costly than the lower denomination ones due to the requirement for increased security features and size. Remember that the Aeris technology works on both paper and polymer notes, so the potential is not hampered. As it stands, banknote sorting machines check notes to see whether they are 'fit' or 'unfit' to re-enter circulation - this is a process that can happen at up to 40 notes per second. Therefore, the process follows that notes with no tears and a certain level (or above) of soiling are sent for cleaning whilst those that are 'fit' do not require it. Notes with tears or more serious problems are discarded. After the unfit notes are cleaned, they are re-tested and if now 'fit' will not be discarded. Those that are still unfit will be sent for shredding. As per the photo above, data provided by Spectra (alongside graphical data) show that some aspects of the note are returned to near uncirculated levels of cleanliness. I would wait until testing results before adding to any holding for detailed 'bank' responses.

Importantly, the majority of notes are deemed to be 'unfit' as a result of soiling rather than tears - soiling often occurs as a result of sebum from fingers that cases the note to turn yellow.  That is all good, but it needs to make financial sense for the central banks to adopt the technology. Using publicly available figures and figures from the company, adoption does actually make compelling reading. I consider two scenarios below.

Scenario 1: US documents estimate it costs circa 5.2 cents to create a $1 bill, and incrementally higher costs for higher denominations of notes. At the lower end of the scale, Aeris is capable of processing 100,000 bills per cycle of circa 2 hours. Therefore, using a simple calculation, to be economical, Aeris has to operate at a cost (during those ~2 hours) of $0.052 * 100,000 = $5,200 or lower. If 25% of the cleaned banknotes are still rejected, the cost needs to be below $3,900 as the cost of replacing 25,000 notes is $1,300.

Scenario 2: US documents estimate it costs circa 13 cents to create a new design $100 bill. At the higher end of the scale, the larger Cool Clean design of Aeris is capable of processing 150,000 bills per cycle of circa 2 hours. Therefore, to be economical, Aeris has to operate at a cost of $0.13 * 150,000 = $19,500 or lower. If 25% of the cleaned banknotes are still rejected, the cost needs to be below $14,625 as the cost of replacing 37,500 notes is $4,875. Once again, the operational costs do not actually change very much as the labour required is fixed, so the economics are improved. 

Taking an all in approach and using the US as an example, the dominant cost of operating Aeris is actually labour, since electricity is available at $0.15/kwatt-hr, and the labour cost is likely to be at around $0.0005/note in the US. Furthermore, the Aeris system does not require to be watched whilst it is working - after all, you would not watch a dishwasher in action. Therefore, there is a clear benefit from cleaning versus direct replacement even at the lower end of the scale, especially since the bank notes lifespan is not shortened through mechanical weakness thereafter. One of the other costs is the cost of servicing the machine, but that is amortised over x > 100,000,000 notes annually and is included in the calculations. This tells us that the higher the denomination of note (i.e. notes with higher creation costs), the more economically attractive it is. Scenario 2 is considerably more attractive than scenario 1, but both make economic sense for a customer. The sheer margin of benefit from cleaning versus direct replacement is the major point.

Financials

Several developments have been made on this front since the initial review, including the release of Spectra's final results for 2014. It does make sense to cover points not included in the results, and which I have therefore adjusted for, beforehand. The first two points are costs associated with business development. Approximately $1.4m needs to be removed from the Final Results cash figure to account for the acquisition of Inksure Technologies, and a further $0.4m needs to be removed to account for the cost of the Aeris creation contract with Cool Clean.

The second cost actually relates to one of the deterrents that stood at the time of the original review - an outstanding lawsuit with a third party who Spectra also had a contract with that had several years left. The company had mentioned that an early and fair settlement would be preferred after the total potential damages costs were reduced in 2013 to $1.27m. Spectra settled the lawsuit out of court and terminated the associated business agreement in order to prevent a potential loss of case and ongoing cash burn associated with the legal case. Naturally this came at an additional cost and the settlement sum totalled $2.1m, which was materially above the reduced damages cost - a $1.8m exceptional charge was backdated to the 2013 results, so will not impact upon the 2014 results and may actually result in softer comparables. The company put forward the point that the legal case plus damages could have cost in excess of the figure paid, and that there would be "significant savings" associated with terminating the contract early. It is sensible to strip out the full cost of $2.1m, taking the total cash to strip out to $3.9m.

Positively though, Spectra was in such an incredibly strong position beforehand that it was able to withstand those charges with relative ease. At this point it is key to recognise that Spectra is not a short-term earnings play. The current valuation is not supported by earnings alone - rather it is supported by a mixture of earnings and the balance sheet, with future potential hardly included.

Looking at the income statement, the 2013 results were pleasing at the top-line as major orders filtered through to a better performance:
- Product revenues grew to $7.96m from $6.4m
- Service revenues grew to $3.15m from $2.47m
- Royalty revenues slightly lower at $0.46m versus $0.49m
- Sustainably high margins across the business of 48%. Flat year-on-year
- Loss of $1.4m pre-exceptional charges
- R&D higher at $2.85m, likely as a result of Aeris. However, this underscores that Spectra would be attractive to a potential suitor that already has R&D. Strip out the R&D cost duplication and the business is appreciably profitable

The first four points are important to comment upon. Total revenues can vary significantly based on the timing of the large product orders so the revenues do not follow a smooth trajectory, albeit there is a stable underlying business that, for example, includes phosphor sales and whilst only generating circa $1m/year has 90% margins. The new manufacturing facility will assist in further boosting margins during 2015 and it's pleasing to see the growth in service revenues as those are more recurring that product revenues. The high margins are testament to Spectra's customer relationships - the banknote industry has high barriers to entry because of the stringently controlled processes and requirement for validation from the central banks.

It remains the case that the balance sheet is excellent. All figures have been adjusted for cash outflows of $3.9m as noted earlier and account for the removal of the court case liabilities that were settled. Cash balances remain substantial at circa $9.5m unrestricted and $2.5m restricted - that equates to £7.2m, and still gives Spectra a low enterprise value of just over £5m. Of course, you can strip the restricted cash out of that, but there is little reason to at the moment. Other current assets total $5.8m. Spectra is a rare case where I would actually attribute value to intangibles because the company has over 100 patents and those will definitely carry intrinsic value. I have therefore stripped out the 50% of intangibles relating to customer relationships and goodwill whilst noting that the customer relationships almost certainly do have value. There are thus adjusted current assets of $24.8m or $23.3m with half the intangibles stripped out. Adjusted liabilities total $5.7m. Using the 50% intangibles figure, I derive a net 'fair' asset value of $17.6m, or £10.6m. Needless to say, that is excellent as it underpins all bar £2m of the current share price and proves that Spectra is far from expensive. Indeed, the balance sheet excludes the potential of both Aeris and the brand authentication divisions.

Although the forecasts for 2014 have been revised down by broker WH Ireland over the period, there is still a residual earnings play, albeit it is not entirely compelling alone. 1.12p in EPS are forecast on pre-tax profits of £0.83m. Spectra has a substantial bank of tax losses to set against profits for the foreseeable future - they actually make Spectra more attractive for a potential suitor, although that is not a desired exit route at the moment. The EPS puts Spectra on a PER of 25, which is very high, but it's important to look past that given the strength of the balance sheet and Aeris' potential. There are plenty of companies in similar positions, where there is a prospective technology, but who do not have any positive earnings, let alone a comparably strong balance sheet. WH Ireland have retained a target price of 40p, which is 43% above current levels and does not really account for the potential that Aeris carries.

"Poised to Capitalise" on Aeris' Potential


Although there remains much to be done, Spectra have noted that they are 'poised to capitalise' on Aeris' potential and that is a sound vote of confidence in the company's future. The balance sheet is still excellent, positive developments have been made since the initial review and Aeris could transform Spectra from a small cap company to a much larger size if the company positions themselves correctly. Regardless, the downside still seems to be limited and, once again, there are numerous companies with less prospective inventions that are valued at significantly higher premiums to their asset values. Spectra is not just about Aeris either, although that is where the most potential currently is - the brand protection is a potential growth driver and there is the underlying business to fall back on. Investors should take comfort from the large number of high calibre institutions who still hold a stake in Spectra, and who were willing to invest at a materially higher market capitalisation back in 2011 when several of the more recent developments were not known. It would not be surprising to see the market attribute a significantly higher valuation to Spectra over the coming months and running into 2015 if they continue to demonstrate tangible progress in any of the divisions. I retain my previous Buy rating.

UPDATE (17/10/14) - Unfortunately the Aeris trials with Banco de Mexico were unsuccessful, and that has prompted me to move Spectra to a No Rating tag in light of the wider market weakness. The market cap remains underpinned by assets, but the timescales on Aeris commercialisation do appear to be slipping, and that could be related to background test results. The move is largely down to the precautionary stance that I am currently adopting. The long-term investment case remains largely intact, but the medium-term one has now been pushed back it appears. However, given the asset backing, I would not 'sell at  any price'.

12 comments:

  1. Brilliant review el1te :0)!
    The bit I found most interesting were the costs. If 0.5 cents a note is the dominant cost then surely the total cost has to be less than 1 cent, perhaps 0.9 cents :0). So there is a giant margin of safety and obviously better for central banks :0). Fab call!

    ReplyDelete
  2. Nice piece; Thanks for all the details and it looks cheap to me although that is obviously a value judgement

    ReplyDelete
  3. Great update el1te. I look forward to Spectra getting the patent which could cause an aggressive rerate. The management are very confident of getting the patent as patents at the moment are about bacterial cleaning mostly. Spectras process uses the supercritical c02 but with changes for banknote so like added substances

    Andrew

    ReplyDelete
  4. Enterprise value of £5m. Looks a steal to me,..

    CraigJ

    ReplyDelete
  5. Great write as usual El1te, many thanks.

    ReplyDelete
  6. Great research as ever, El1te.

    A couple of questions on this company:

    - Are you aware which two G8 central banks they have business with? Why the need for secrecy in this respect?

    - While central banks are clearly slow to release information around their operating expenses, I found a study which showed that the US Fed spent over $850m replacing bank notes in 2008. Did you find anything further in terms of what % of this cost related to replacement of soiled (as opposed to torn, etc) banknotes?

    - Is the 75% cleaning success rate you use in your illustative calcs above based on company data on testing? Clearly this rate will be a key factor in proving the technology.

    James

    ReplyDelete
    Replies
    1. Hi James,

      1. Can't tell the G8 central banks at the moment. Government contracts especially with relation are often very difficult to track down - suspect that the lack of a name is just down to confidentiality. Should really refer to it as the G7 now after Russia was kicked out!

      2. It's estimated that circa 60% to 80% of all rejected notes are from soiling issues, and one study (unreferenced) attributed 81% of rejection to soiling, 9% from tears (and associated), circa 5% from graffiti etc. In essence, the vast majority are removed due to soiling.

      3. A 25% post-cleaning rejection rate is a sensible ballpark estimate - checked that with NL. I.e. One in four notes post-cleaning are still too soiled to be re-used and the other three can be cleaned to standards acceptable for re-use. It's interesting though. Just for arguments sake, if there was a even more conservative 35% post-cleaning rejection rate for $1 bills and using a 100,000 vessel, the 2 hour process would need to operate at a cost of under $3380. That is 3.4 cents per bill. Once again, we're looking at less than 1 cent for cleaning so still very much an attractive option, even with the lowest denomination note. Higher denomination note economics are even more attractive. The same still applies for a 50% rejection rate, although that's almost certainly far too high.

      In lower economically developed countries, labour costs (i.e. the dominant cost) is likely to be lower, meaning that the fixed element is lower and OPEX for Aeris is lower.

      El1te

      Delete
    2. Thanks for the reply, El1te. Using those rates it seems this solution would certainly make economic sense for a central bank to adopt.

      It appears that the most important aspect to consider at this point are the integrity of the IP (i.e. patent being granted), given the technology would appear to be relatively easy for a competitor (or central bank!) to replicate.

      James

      Delete
  7. Price hit pretty bad today

    ReplyDelete
  8. Given your access to Dr Lawandy, can you get any more detailed information as to exactly why the BoM tests were unsuccessful, given the apparently perfect results that were previously obtained for other notes. Is it just a case of BoM saying they can't afford it right now and want to carry on with their pet plastic notes project.

    ReplyDelete
    Replies
    1. Hi,

      I recommend you email him at nlawandy@spsy.com; he is usually prompt in replying. Whilst the share price reaction was no doubt very harsh based on the news released, sentiment across the market is fragile at present and will take time to repair

      Best,
      El1te

      Delete
  9. Hi
    As at the date of this note, 20/11/14, the case for investing in this company has got even stronger IMHO.
    Market cap = £9m and SP = 20p. Based on the 30/6/14 interims, the numbers are :
    • Revenue 225% ahead of previous year
    • Gone from loss making to a projected FY £589k profit
    • Forward PE = 15.4
    • Stripping out cash of £6.1m. PE drops to 4.9
    • Net Debt = -£3.1m, making EV = £5.9m
    • Net tangible asset value = £11.1m, greater than market cap of £9m
    • Bullet proof balance sheet
    What has changed of course is failure of the Bank of Mexico Aeris trial. So to me, even without Aeris, this is an attractive proposition. With Aeris, and assuming it achieves a fraction of what it can potentially, this share is likely to rocket.
    Needless to say that I am long.
    All E & OE, DYOR, etc.

    ReplyDelete