Beeks Financial Cloud (BKS): AWS of Finance, or a commodity compute provider?

Hello,

I have started a very small position in Beeks Financial Cloud, but would like to discuss the company. Open to criticism of my thesis. The company currently trades at a rich market cap of 59.46M at a price of 118p. While I feel this is too high at 6.4x last year’s revenue of 9.3M, I feel the company does operate in a interesting space and if successful could be like the AWS of FX and futures trading markets.

Beeks Financial Cloud is a cloud computing & infrastructure provider for the financial services industry. It was founded by current CEO Gordon McArthur in 2010 and provides virtual private servers with direct, cross connects (fibre optic cable) into financial exchanges. The Company is focused on supporting automated traders in the foreign currency and futures markets. The value proposition they provide is server infrastructure for direct connectivity with low latency to financial trading exchanges. In automated trading, milliseconds matter and a direct cross connects connection is vital for many machine driven trading strategies. Beeks’s clients include IG, Forex Capital markets, Playtec PLC and various other hedge funds and market participants. Beeks has 18 datacenters. Some of the prominent ones are in Singapore, London, NYC, Paris, and Sydney. They recently acquired Velocimetrics which provides trade analytics software. I believe adding analytics software is smart because it enables you to gain more wallet share from existing customers and should usher in higher GM%.

Gordon is an owner operator who owns 53.63% of the company per the latest Annual Report listed below:

Revenue grew 27% to 9.36M, though it is important to note that the rate of revenue growth has decelerated every year since the company floated in 2017 (revenue grew 41% to 5.58M in 2018, 32% to 7.32M in 2019). Gross margins have been around the 50% range (good but not great, as the business is an infrastructure business, not a SaaS business). EPS was 2.37p in the most recent financial year. Earnings are not impressive as the company has had to invest a lot in the equipment at the 18 datacenters. As a result ROCE will not be impressive, at 7.9% based on the last annual report. The company has approximately 0.75M in net debt has been used to acquire Velocimetrics and pay a dividend. I do question the wisdom of paying a dividend, as the company has mostly been in investment mode over the last 3 years. This is evidenced by the lack of meaningful free cash flow. Will monitor how management views capital allocation going forward.

The silver lining is that management has mentioned the buildout of datacenter capacity is mostly complete. The company has a footprint in North America Europe, and Asia, but does not have any coverage in Africa or South America. If the buildout is mostly complete, the company may benefit from higher margins due to the Velocimetrics software. In addition, the company has recently signed on two “Tier 1” clients. (these are not mentioned, but we must assume they are very large banks/institutions). The pandemic has posed a challenge to the company as to win large clients the company must perform and present its value proposition directly with the client. The goal for management is to sign up more Tier 1 clients and get in the door for more infrastructure outsourcing discussions.

The bull case is that growth continues and due to operating leverage more revenue hits the bottom line with a decrease in investment spending. The business is a capex intensive business to get into, and one must have a deal with the financial exchanges to be co-located to have direct access to the exchanges. How many companies can get this?

The bear case is that the company must continue to perennially invest in the business to fend of competitors. Can Amazon Web Services, Google Compute, and Azure get into this space? Management did mention on a 2018 Mello Derby day that customers who did switch to an AWS usually switch back to Beeks (see below link to the Mello Derby Day video from Piworld, discussion of compeition is 21:50):

https://www.youtube.com/watch?v=5H6F8U3cJ1w

The company also competes against Fixnetix (owned by DXC who outsources bank’s infrastructure needs) and Lucera (owned by Cantor Fitzgerald). Both competitors are impressive, however DXC has faced its own management challenges in its acquisition of the HPE Enterprise Services business made a few years ago. A ctrl+f search on the DXC annual report yields no result for Fixnetix.

From a Regulatory perspective, some folks may worry about regulation of high frequency trading. Gordon addressed this in the Mello Derby day. He mentioned it was not scalable to do HFT on a shared platform like Beeks and that Beeks’s focus is automated trading, not HFT. Not sure how to view this. Governments are capricious, time will tell if this is an issue.

One accounting risk to watch is the intangible assets related to acquisitions like Velocimetrics. I will be watching to see how the deferred consideration pans out for the acquisition. There is approximately $1.8M in Goodwill related to this transaction and the discount rate for the good will was set at 29.5%. I will be scrutinizing future reports to see if this discount rates creep down to maintain the goodwill, or if any impairments occur.

In terms of longevity of demand, I don’t question that Hedge Funds, and banks will still be around decades from now. Human nature and the desire to play the markets will not disappear. The question is whether Beeks can serve as the computing platform for this behavior and sign on more Tier1 Banks and scale the business while controlling costs.

1 Like

Hi Anish,

Many thanks for the write-up. Not looked at BKS before. As well as the Mello presentation, BKS has presented on the InvestorMeetCompany platform, where management provides further explanation of the ‘moat’ (e.g. “Q8: Can you please reiterate what you see as your defensible moat (this is a fairly crowded marketplace - what sets you apart)?)

Valuation is an obvious drawback. You are paying for a lot of expected progress. Doubling up the latest interim underlying EPS gives 1.9p per share = 63x at 120p. But yes the company has notable attractions. Recurring revenue, upbeat trading, etc. Alongside his large shareholding, the chief exec is relatively young (mid-40s) and takes only a £60k salary – both positives in my view.

So far the economies of scale have yet to occur, at least on simple measures such as revenue per employee and revenue per computer equipment:

2017 2018 2019 2020
Revenue (£k) 3,970 5,583 7,352 9,360
Employees 17 23 29 41
Revenue/Employees (£k) 234 243 254 228
Book cost of computer equipment (£k) 2,219 3,619 4,839 7,590
Revenue/Book cost of computer equipment 1.79x 1.54x 1.52x 1.23x

Recent computer expenditure in particular has not created the revenue levels of the past, but that may be due to the new expenditure having yet to ‘mature’. You could perhaps justify the rich rating with proven economies of scale, but future scalability presently has to be taken on trust – which is not ideal.

I note in the 2020 report that the expected useful life of the computers was extended from 3-4 years to 5. That is quite important at as, in theory, BKS is extending the replacement time cycle by 43% (from 3.5yrs to 5), which should equate to less capex over time.

BKS explained:“Following this review, the Group considers the estimated useful life of its computer equipment to be 5 years. The Group believes that this provides more reliable and relevant information to the users of its financial statements with regards to the length of time economic benefits are consumed over

I could not find any evidence of BKS disposing of PPE at a profit over its book value (i.e. the computer assets had been ‘over-depreciated’)

But for comparison Amazon’s 2020 annual report says:

We review the useful lives of equipment on an ongoing basis, and effective January 1, 2020 we changed our estimate of the useful life for our servers from three years to four years. The longer useful life is due to continuous improvements in our hardware, software, and data center designs.

And Alphabet’s report says:

“In January 2021, we completed an assessment of the useful lives of our servers and network equipment and determined we should adjust the estimated useful life of our servers from three years to four years and the estimated useful life of certain network equipment from three years to five years.”

So cloud servers are indeed becoming more durable, so perhaps future capex maybe not be as great as before and therefore help the missing economies of scale to emerge.

Maynard

2 Likes

Thanks for pointing out the useful life changes. I will definitely watch that to make sure they are not gaming the P&L. The company needs to continue adding Tier1 clients. I think the move to cloud will continue, the question will be if Beeks can out compete fixnetix and lucera. I do like Gordon as a leader, seems like a no nonsense CEO, but we will have to see what happens.

Anish

Beeks Financial Cloud just did a 5M pound placing to fund a build out of their private cloud business and potential future acquisitions. The investment phase is not over and looks set to continue for Beeks. It doesn’t look like profitability will increase in the near term. While it is good to invest in the product, I am disappointed Beeks had a dividend in the first place knowing that it needed more investment. Gordon is now owns 48% of the company instead of 53%. We shall see what happens and how they allocate the money.

Hi Anish

Agreed. The RNS revealed the expected use of the proceeds to be:

· approximately £3m for investment to support the launch of the Beeks Private Cloud offering;

· approximately £1m for the repayment of the Group’s RCF facility; and

· up to approximately £1m for additional working capital, including for the evaluation of M&A opportunities.

So only 60% of the proceeds goes towards true ‘expansionary’ investment. Not ideal.

The RCF facility was created at the time of the Velocimetrics deal and the latest interim statement said the facility had switched to Barclays to secure better terms. Why extra shares are being issued to reduce the RCF is a mystery. Further M&A looks on the cards, too, which I would say is not great for what the 60x P/E would suggest is a dynamic, organic-led growth company.

The latest interims also contained this smallprint:

“In the current period, the fair value of the final contingent consideration has been reassessed. Noting that a modest delay in new business wins or delivery in the next six month period could result in a substantial change in the amount due under the earn out terms. The directors believe that the probability of achieving the maximum earn out target for the full financial year end to 30th June 2021 is not as likely and have reduced the provision to represent the consideration paid based on a lower earn out target being achieved. The final consideration has therefore been reduced from £1.96m to £1.36m with the difference £0.60m credited to the profit and loss in the period.”

A 30% reduction to the final earn-out could mean Velocimetrics’s 2021 performance coming in 30% below the expectations at the time of the deal. Again, not ideal.

True, although scrapping the £180k dividend may not have relieved BKS of needing £5m. I still get the impression the economies of scale and free cash generation from the existing network/servers have yet to materialise.

Maynard