Weekly Commentary 02/08/21: Credit and Contagion

The FTSE 100 bobbled about just above 7,000 last week. Nasdaq 100 was down less than half a percent, as strong results from Alphabet, Microsoft and Apple all beat analyst forecasts on the top line. This was offset somewhat by Amazon, which reported sales growth slowing as lockdown measures eased. Maybe Jeff Bezos has timed his departure well? The FTSE China 50 index was down -9% last week, and is now down -15% since the start of the year. Most of the damage has been done since the start of June.

China Evergrande’s shares fell another -12% early last week, as the Chinese property developer reversed a decision to pay a special dividend. Then following a “dead cat bounce” the shares fell another -10% in the first 15 minutes of trading on Friday morning. This looks like a “death spiral” for Evergrande shareholders; normally when a company’s bonds are trading at 50c in the dollar, as Evergrande’s bonds are, it implies that shareholders will be wiped out. The credit rating agencies have been downgrading to junk status. The group is huge, over $300bn of liabilities and 231 million square meters of land in 234 Chinese cities. As well as bank loans and bonds, the developer has relied on commercial paper and pre-sales of flats to customers to fund growth. The city of Shaoyang ordered the company to halt pre-sales last week, before saying the restrictions would be reversed. This does look like the Chinese authorities trying to prevent systemic problems, but the current Chinese credit cycle is unsustainable; more of the same will be different.

Beijing also suggested that Chinese online tutoring companies should not make a profit, which caused a sell-off in Nasdaq. There’s a related worry that the Cayman Island structure that corporate financiers have used to circumvent Chinese foreign investor restrictions (Variable Interest Entities or VIEs) allowing Chinese companies to be listed in New York could be next. The aggregate value of VIEs like Alibaba (down -14% in the last 3 months) and Tencent (down -22% in the last 3 months) is around $2 trillion in market cap. Maybe all the problems are contained within the Chinese border. But I think it more likely the contagion spreads, and this is what drives investors narrative in H2 rather than stories about Covid.

Other signs that make me nervous are:

i) Robinhood’s IPO. Rather than first day drop of -8% last week, I think the timing of the profile financial IPOs can indicate that we’re close to the top of the cycle. For instance the private equity firm Blackstone’s IPO in June 2007 just as the last credit bubble burst, and before that Blackrock, the fund manager’s IPO, Oct 1999 at the top of the internet boom. The average “first day IPO pop” has been +39% in the US in 2021.

ii) Bank results. Deutsche Bank has reported better than expected revenue growth in Q2 and in last Friday’s NatWest H1 results, they announced a 3p dividend and buying back £750m of their own shares. Bankers’ optimism tends to be greatest just before the cycle turns.

There’s a fascinating podcast interview with Russell Napier, about the 97-98 Asian crisis, which started in mid-1997 with the Thai Baht being devalued and ended with the bailout of hedge fund LTCM. He points out how slowly financial contagion spreads, so for instance in the summer of 1997 people thought Thailand and Indonesia were in trouble, but more developed north Asian countries with trade surpluses and healthy foreign exchange reserves like Taiwan and Korea were immune. It took 3-4 months for the pressure to build on those countries, but eventually their currencies, which had been pegged to the dollar, devalued too. Investors thought Russia’s fixed exchange rate might come under pressure, but the big shock was caused more than a year later by Yeltsin defaulting on the rouble-denominated debt in August 1998. LTCM, the hedge fund, was hardly involved in emerging markets and mainly exposed to long dated US government bonds and derivatives, but the knock-on effects meant it failed in September 1998. The contagion was caused by unwinding the carry trade and a slow realisation that leveraged losses in one market caused forced selling elsewhere, which led investors to hasten the unwinding of their positions. Feedback loops which had been virtuous circles turned vicious.

The same was true a decade later. In January 2007 HSBC warned on profits due to subprime lending exposure, Northern Rock imploded six months later in June 2007, but it took more than a year for Lehman Brothers to fail in September 2008 and the stock market bottomed another six months later in March 2009.

This is odd; information travels at the speed of light, while financial contagion and repricing of risk spreads much more slowly. I think that’s because credit crises are not discrete independent events, they are chaotic processes, with feedback loops. The same information is available everywhere, but it still takes time for our brains to make the connections and see what is relevant. The best way to prepare for a crisis is not to be overly leveraged (unlike LTCM which had 10,000 swaps with notional value of $1.25 trillion on its books) and have some spare cash ready to deploy.

This week I look at Games Workshop’s FY results to May, Sylvania Platinum’s disappointing Q4 update and K3 Business Technology, which could be an interesting turnaround situation. All three of these companies have net cash, which gives considerable comfort even if sales or profits disappoint.

Sylvania Platinum Q4 to June 2021

This South African platinum company announced a Q4 update to June. Rather than mining platinum they operate Sylvania Dump Operations which takes chrome tailings from host mines and recycles into Platinum Group Metals (4E is platinum, palladium, rhodium and gold). The FY Annual Report will be published in September, this was just a Q4 update with headline figures.

Revenue down Revenue fell -35% to $48.4m as their 4E gross basket price fell -11% to $4,059/ounce on the previous quarter. Rhodium was down -19% in the quarter, which I think many investors have been following and so the 4E basket price fall should not have come as a surprise. However the -35% decline in revenue does look disappointing. The rest of the revenue fall was caused by the “sales adjustment” down -98% versus $15m in Q3. This adjustment is calculated from the difference in basket price for the period between delivery and invoiced for several months later. The company are not particularly good at explaining the mechanics of how this figure is calculated.

In the quarter they produced 16,289 4E PGM ounces, or 70,043 ounces for the year as a whole, in line with guidance. The Q4 production figure was down -6% vs Q3, which management attribute to a significant increase of open cast chrome tailings; these have oxidised and hence generate lower recoveries than from chrome tailings sourced from pit mines. It sounds like there is a trade-off between supply of higher quality tailings and volumes. Management expect PGM recovery to improve to 52-54% range (vs 51.2% reported this quarter) in FY to June 2022, so we should see better recoveries in the coming year.

Costs higher Cash costs rose +16% in rand terms and +22% in US dollar terms vs Q3 this year, which is also blamed on the higher costs of securing Run of Mine (RoM) tailings and also a +15% increase in electricity costs. The higher RoM costs were due to chrome mines shutting down in response to falling commodity prices last year, so it’s a little disturbing that supply from host mines is still an issue in the most recent April to June quarter. Plus there was $2.4m spent on capex during the quarter.

The story with SLP has always been that although PGM mining is a commodity industry, the dump operation’s fine grinding mills and flotation circuits are lower cost than digging PGMs out of the ground. Cost in Q4 were $912 per ounce (Q3 $745 per ounce) and those higher costs were the likely reason that the shares fell -15% on the morning of the results, because higher costs reduce the structural advantage that SLP enjoys versus the other platinum miners.

Cash The company has $101m of cash on their balance sheet at the year end, down by $1m over the quarter due to paying a $35m increased tax bill (the South African Govt is short on money, and has introduced a mining royalty charge as well as normal income tax) plus a windfall dividend of $14m was also paid during the quarter. I’ll wait for the FY Annual Report before forming a view on cashflow, but in previous reporting periods I have been concerned about the growing trade receivables (ie customers might be struggling to pay SLP because of the spike in rhodium) and the variance of the “sales adjustment”.

Valuation That $101m of cash represents a quarter of the market cap. Quarterly profits now reported for FY 2021 sum to $96.5m. Deducting cash from the market cap, and putting the shares on a FY June 2021 multiple puts the shares on less than 3x historic earnings. That figure rises to 7x historic earnings if you use last year’s FY June 2020 figure, which included placing operations in “care and maintenance” for 6 weeks.

Opinion Such a steep -35% revenue decline combined with “all in sustaining costs” +34% in $/oz is clearly disappointing. No doubt civil unrest in South Africa hasn’t helped the investment case either, though management point out that their operations are in Mpumalanga and North West province where no riots have occurred. For me though, the key question is whether SLP can maintain their cost advantage relative to the other platinum miners, rather than short term spikes and crashes in rhodium. I’m hoping that this is just one bad quarter when various negative influences came together. I own the shares, but I’m a little less confident than I was before the results.

K3 Business Technology H1 results to May 2021

This technology firm which helps retailers with online sales and ERP and has IKEA as a customer, released H1 results to 31 May. NB It’s very easy to confuse it with K3 Capital (ticker K3C) the professional services firm, which has a FY to May.

K3 Business Technology (ticker KBT) reported flat revenues of £23m and a loss from continuing operations of £4.3m. That seems disappointing for a technology company that helps retailers, for instance by implementing a “click and collect” system. Presumably many of their customers have struggled with lockdowns, supply chain issues etc and are fighting fires, but hopefully this means the outlook for future investment spend will be to KBT’s benefit? Or are they serving a sector that has been “Amazon’ed”, the equivalent of blacksmiths still making horseshoes for horses when Henry Ford’s Model T are rolling off the production line?

Earlier this year they sold a non-core unit, Starcom Technologies, for £15m in cash. Hence the balance sheet has swung from a net debt position last year of £7.4m to net cash of £4.4m, which is obviously more comfortable for a firm reporting ongoing losses. The cash position was also helped by the conversion of a £3m shareholder loan from Kestrel Partners (who are represented on the board by Non Exec Oliver Scott) and Fastigheter AB (who are represented on the board by Non Exec Johan Claesson) which converted at 168p, the market price on 26th March this year.

Both gross margin 56% and recurring revenue 69% fell slightly y-o-y, though both numbers are still attractive. I’m struggling to understand the company’s cashflow disclosure. The £10.3m cash benefit from the sale of the subsidiary in the cashflow statement, doesn’t match the £14.7m or £13.8m of cash claimed in the headlines at the top of the RNS. There’s a large working capital negative item of -£9.8m from a reduction in payables. The commentary blames half of this amount on a working capital adjustment of £5.1m for discontinued activities. I think the £5.1m adjustment should be offset against the £14.7m in cash they sold the business for, otherwise the headline figure seems to be a rather flattering way of presenting the transaction, in my view.

A further reduction to cash from operations is a £3.5m tax payment in the UK and Netherlands which reduces cash from operations. K3 has reported £37m of cumulative losses since FY 2017, so it’s rather a shame that these losses can’t be offset against tax. When I’ve asked companies about this in the past, they say that the tax benefit has to occur in the same jurisdiction that that the losses occurred. I’d guess that’s what’s happening with K3, but there’s no explanation.

Outlook They say that they are confident that there’s a significant market opportunity, building their own IP sales, which are higher margin and recurring revenue. I find the statement vaguely worded which can signal lack of momentum. The Chief Executive Marco Vergani was appointed in March, which probably explains his reluctance to give more detail. This seems a similar situation to ULS Technology, another company with a decent story but significant execution risk and where Kestrel also own a large stake and replaced the Chief Executive with someone whose background they thought would be more suitable to drive a digital business.

Ownership Kestrel Partners own 25%, Richard Griffiths 11%, and Liontrust 10.4%. Per Johan Claesson, a non Exec Director owns 24%. He’s a third generation Swedish corporate financier, who owns this property and investment company. Because of the large insider ownership, the free float (ie percentage in public hands) is just 15.5%.

Forecasts finnCap are the broker, but haven’t published any forecasts. For those readers who don’t know how corporate brokerships work, the company management have likely asked finnCap not to publish forecasts to the market. When I was corporate broker I had to say nice things about Barclays and HBOS and reiterate their guidance with my published forecasts because Credit Suisse were their broker. Then, after I left CSFB and was no longer broker, I could publish whatever I liked, happily writing “sell” notes and openly calling into question the nonsense that bank management were saying to investors.

So it’s unfair to blame finnCap for the lack of forecasts, or for making upbeat comments on the results. Management are paying finnCap a retainer and probably don’t want their broker to publish numbers for fear that they then disappoint the market. You can say that it’s a silly system, but that’s how it works.

Opinion So that leaves us with the shares trading on 2x historic revenue, with a healthy gross margin but poor historic RoCE and EBIT margin. A classic turnaround situation. If the new Chief Exec can reverse several years of losses, then it’s good value, but there’s considerable execution risk. There are a couple of old presentations on PI World from a few years ago. But I’d really want to see the new Chief Executive present and see if what he says makes sense.

Games Workshop FY to May 2021

The fantasy miniatures and gaming company released FY results with revenues +33% on a constant currency basis. Statutory revenues were £353m (previous guidance on 20th May was not less than £350m) and PBT was £151m (previous guidance not less than £150m). Operating profit on a constant currency basis was up +73%. The gross margin improved to 72.7%, which many software companies would be envious of. GAW had £85m of net cash on balance sheet at the end of May. Despite being “in line” with guidance the shares fell -5%, probably because some investors have come to expect a positive surprise.

Voluntary Disclosure Readers will know that I’m interested in company’s “voluntary disclosure” – what management choose to communicate to investors beyond what is required by the accounting standards and reporting obligations. The previous chairman, Tom Kirby used to write opinionated commentary which was engaging to read. The new Chair, Elaine O’Donnell’s prose is less colourful. Chief Exec Kevin Rowntree’s narrative is focused on what has happened and what will drive the business in future. Nevertheless the way management write remains exemplary to other companies. SharePad user @Vivek posted a word count on Twitter: “Customer” appears 107x, “Service” 70x, “Staff” 76x, “Adjusted” 6x and “EBITDA” 0x.

The company’s analyst PowerPoint presentation is a masterpiece of understated simplicity. 14 slides (the last two numbered incorrectly) mainly of text, with just 4 slides with single graphs (sales by channel, sales by geography, core business operation profit trend, return on capital trend – see next page).

Games Workshop Return on Capital % (slide 8 of 14)

Presumably only a company that reports 184% Return on Capital could present something so uncomplicated. For signalling, GAW’s presentation rivals only the Berkshire Hathaway website, which looks like it was built for free by a teenager who was trying to learn HTML in 1997, and the design hasn’t been updated since then.

Ownership Baillie Gifford own 7.3%, BlackRock 6.9%, Schroders 6.5%, Abrdn 5.2%, JP Morgan AM 4.7%, Vanguard 4.1%.

Outlook They say that they have internal plans for future performance which they don’t wish to share with the market. SharePad shows forecast revenue growth of +9% for this year and next with EPS of 387p FY May 2022F and 417p FY May 2023F, putting the shares on 29x and 27x respectively. GAW tends to generate free cash, so the price to FCF multiple is perhaps a better measure 26x May 2022F and 22x May 2023F.

Opinion Forecasting growth for this company is difficult, revenue flat lined for 6 years between 2010 and 2016. It then trebled from £118m in 2016 to £353m just reported FY May 2021. They are an international business with 77% of sales coming from outside the UK. But management are investing to grow further, they have hired in-house Chinese and Russian translation teams who will be in place in H1, and they plan Warhammer cafes in Shanghai and Tokyo. They also launched nine new computer games and have several planned for next year, including Warhammer 40,000: Chaos Gate, developed in partnership with Frontier Developments. I own the shares, and I know that it’s a favourite among readers too.

Bruce Packard


The author owns shares in Sylvania Platinum and Games Workshop

Leave a Reply

Your email address will not be published. Required fields are marked *