April 2023 Investment Update

Dear investors and well-wishers,

The fund returned 5% in March and 20% for the quarter.

These are our top 10 positions:

Camplify, our largest position in Australian small caps, just reported 120% organic revenue growth year-on-year, forward bookings up 114%, and its second consecutive quarter of positive free cash flow.

Light blue bars represents Camplify’s recent acquisition of Europe-based PaulCamper

The biggest question in software investing

How will the step change improvement in artificial intelligence affect the performance of software companies?

Developers are now vastly more capable, and soon entire codebases will be queryable by large language models in ways that exceed the most experienced developers. Engineers will soon be able to ask a computer why things were done the way they were, and how best to improve them.

At the moment it seems this could go in two directions.

Developers (and other employees) could become so much more productive they become more valuable to companies, which along with competitive pressures could lead cost structures to stay roughly the same, though with a much faster rate of development.

Another possibility is that companies become vastly more cost efficient, right at the time when most listed software companies are pushing hard in that direction.

The opportunity may be largest for companies that have stacked years of cohort revenue and might now be able to harvest those cash flows on a much reduced cost base. This includes many listed companies trading at decade valuation lows down 65-80% from the highs of 2021.

But it’s not yet clear. Competition will certainly intensify across the board, and tedious tasks like migrating solutions will become far more manageable once the heavy-lifting is done by artificial intelligence. This could erode the competitive advantage and ‘stickiness’ of existing software, so it’s a threat and an opportunity for everyone.

Another possibility is that companies build more in house and cheaply replicate much of what they are paying through the nose for currently.

We can say two things with confidence:

  1. Much more software will be written
  2. Compute will dramatically increase

The first part of this – investing in semiconductors – has, so far, played out. The second part, so far, has not, with companies like MongoDB and Digital Ocean still under pressure in the market. Last night’s steady results from Microsoft Azure and Google Cloud may steady things.

Digital Ocean is particularly interesting with a $500 million buyback authorised against a $2.7 billion enterprise value, and a compelling cost-effective hosting alternative when tech decision-makers are more cost-conscious than ever.

Digital Ocean EV/Sales and trailing 12 month revenue – there are plenty of companies that look like this. 

AI aside, we’re finally starting to see large efficiency gains across the software space, with some companies improving GAAP margins by 20% year-on-year, with similar moves expected over the next twelve months.

This continues the parallel with the travails of the energy industry several years ago, when a spectacular bust led to a drive to efficiency and shareholder friendly capital management that laid the path for the extraordinary recovery from 2020 to today.

We can see this elsewhere too.

We wrote in our last letter that Sea Limited made a profit of $420 million in the fourth quarter up from a ~$600 million loss the year before, a turnaround in quarterly profits of over $1 billion. And they are continuing to increase their take rate.

I’ve wondered what the Amazon of this cycle will be. Sea Ltd is a good candidate. The stock dropped 90% despite the company growing throughout. Sea completed the full circle from relative unknown (when we bought) to the hottest stock in South East Asia, all the way back down again. The firm has billions of dollars of cash, and is now generating sufficient PnL that largely squashes the bear case. After a strong rally the stock is still down over 80%.

Meanwhile profitable, growing small caps have continued to perform. Paying 5x earnings for Crocs has so far turned out well.

I have noticed however, that the insider buying (which was one of the things that put us on to this company at the lows mid last year) has now switched to selling.

Signs of life

There are signs of life elsewhere. We had a few small biotech companies that have finally bounced off the lows. These are small positions that in some cases we have held for quite some time.

Cyclopharm is a local example. Cyclopharm sells Technegas, an ultrafine dispersion of radioactive Technetium-labelled carbon, used to image lungs.

FDA approval has been elusive, despite success in 64 countries around the world. We judged that if Technegas was good enough for the European Medicines Agency, it would eventually prove good enough for the FDA. A US approval would give a significant boost to Technegas’s existing $23 million revenues, with a potential market size estimated at $180 million (Cyclopharm’s market cap is ~$200 million).

Schrödinger has also come off the lows. This is one of our largest positions in the life sciences and we discussed this in past letters. The mix of software revenues and a deep, diverse pipeline of partnerships with leading pharmaceutical companies is a compelling business model in such a risky part of the market. After a decent rally the stock is still down 75%.

We also initiated a new investment in NU bank, the fastest growing bank in South America.

Cohort analysis shows mature customers are generating four times the revenue of new customers, and new customer growth itself is strong, suggesting a long runway from these two compounding effects.

Revenue growth came in at 110% over the last year, a period over which the stock dropped ~70%. Nu’s net income margin is 13%, driving a return-on-tangible-equity of 25% with a target range of 35-40%, supporting a price-to-book ratio of 4x.

There’s a similar story in Crowdstrike, where management is confident they can quadruple revenues off existing customers alone.

The biggest risk to Crowdstrike is competition from Microsoft, which in the past, kept us on the fence. When tested head-to-head, Crowdstrike has a decisive upper hand, but we can be sure that Microsoft is investing heavily to close that gap.

On the positive side, there’s a principle in security that you don’t want the same company delivering core software to be in charge of protecting it, as if a malevolent finds a security gap, that can be particularly devastating, as there is no other line of defence. We have a small position, this certainly has the growth and customer love (it’s a clear favourite of security professionals), but we want to be sure.


One way we are adapting our strategy is to develop tools that reduce our tail risk – which we’ll be the first to admit was sorely lacking last year.

I’m often asked if we are changing the strategy.

Exceptional organic growth, customer love, market leadership, remain the best ways to identify a surprisingly large amount of the top performing companies for most of the past few decades. Ofcourse there were two harrowing multi-year exceptions, 2000-2002 and 2007-2009. We are now at the later stages of a third period.

But it seems every 10-20 years these kinds of companies are going to suffer >70% drawdowns, and I personally never intend going through one of these again.

I often wondered what it was like for the managers who backed Amazon, for example, seeing how the internet would decisively change the way we purchase and that this would offer a myriad of new business opportunities, that would largely accrue to the largest.  (Amazon is not alone, you could pick any number of leading companies in software, internet or semiconductors at the time, including NVDA, AMD, Priceline, eBay, Adobe, Autodesk, etc all of which dropped 80-95% only to make full recoveries and advance 20x,30x or even >100x in the years that followed).

Risk models are particularly effective for stocks post significant moves in both directions, which just so happens to line up closely with the characteristics of the kinds of companies we invest in.

If you applied this to something like CBA or Wesfarmers the results would be very different and would generally lead to much lower returns.

Risk framework applied to Crocs over the last few years

The same risk framework applied to Crocs since 2006 – a >40x return.

These kinds of tools would have been very effective in companies like Sea Ltd when there were extraordinary rallies and a large crash.

It’s worth noting that the same frameworks that would have given much needed sell signals in late 2021/early 2022 have flipped to buy across our portfolio.

Risk framework applied to Sea Ltd over the last few years

Also astonishingly effective for fast-growing companies that go through multiple boom-bust cycles like semiconductors – this is NVDA through the tech crash, the GFC and 2021-2022

The second major aspect of risk was policy, which proved vastly more relevant to our outcomes than expected, dwarfing fundamentals. Most of our companies have posted steady growth throughout the last few years and have swung from ice cold to red hot and back again multiple times.

We had some big winners in Chinese tech a few years, only for the cancellation of Ant Financial’s IPO to trigger an 80% fall in the Chinese tech index. If there’s a renaissance in those companies, you can be sure it will be led by policy.

We launched in mid 2018 and invested just in time for the first tightening by Powell, triggering one of many 30% drawdowns in small cap stocks.

Powell’s famous pivot that December triggered an extraordinary rally to early 2020. Politicians can’t be blamed for the COVID sell-off (…) but in late March Powell gave a crystal clear buy signal which was matched by models like those above. This lasted with minor interruptions until December 2021, when Fed officials not only signalled radical new tightening, but also sold all of their own stocks personally. Shortly after, risk models flashed red as speaker after speaker emphasised financial conditions – by which they meant they were explicitly targeting lower equity and credit prices.

At this time, fundamentals of many of our core companies were still accelerating, which proved the trap. In fact, most of our companies continued to post record revenues throughout this sorry period.

Markets are reminiscent of forests, where lightning often strikes, but usually only burns down a tree or two. This time tinder had been building and when the strike hit, it caused a bushfire, triggering the unwind of perhaps two decades of growing tech exposure. Companies were cut down 60-90%, even while – without wanting to repeat too much – many were posting record revenues and profits.

So there were two ways we could have minimised last year’s drawdown and held on to the bulk of the hard-won gains in years prior. We could have reduced risk when the Fed turned hostile towards asset prices, and we could have used price-based risk models to reduce our downside as well. Both of these would have kept us fully invested in that period from early 2020 to late 2021, and indeed, throughout most of the bull market from 2009.

Most importantly these methods would have allowed us to maintain most of our market leading performance during those nearly forgotten periods where growing companies went up rather than down in price.

Where are we now? 

The bushfire has burned through: exposures and valuations are at decade lows. Momentum indicators are signalling buys across our space, but there is one piece missing: the Fed is still tightening rolling out speakers asking us to ignore the data we can all see and that they will continue to pursue tighter financial conditions.

But a decisive turn in inflation is around the corner, and we are close to the point where lower prices finally show up in the one number the Fed counts most: the year-on-year CPI print reported in the press and recorded in the history books. As last year’s high inflation months roll out of the series in May and June, the Fed Funds rate could soon be 2-3% higher than the actual year-on-year CPI.

Truflation source

Which would suggest we could be close to a turning point more decisive than those of December 2018 and March 2020.  But it hasn’t quite come yet.

Back to fundamentals, this has been a frustrating period. Wins in companies like NVDA, Crox and Achieve have been counterbalanced by deteriorating breadth where only a handful of companies are driving performance and most are performing poorly. And we anticipated a software recovery which so far has not come, though when it does, the returns from these lows will likely be exceptional.

Another one of those charts. Sometimes these extreme mark lows, sometimes they don’t, presented for interest.  

It hasn’t helped that tech executives themselves don’t seem to see value at current prices.

Recent insider sales by Datadog executives 

What you can expect from us from here is a continued focus on fast-growing companies, of which there are still many, and a renewed focus on risk controls. Our pathway back out of this is a portfolio of fast-growing stocks that are trending up and leading the market in both fundamentals and price. Many fast-growing companies are still trading close to 3 year lows while continuing to stack cohort revenue, right as Silicon Valley has made a revolutionary advance in AI that is already driving a leap forward in productivity and value creation.

We are so close.




The information in this note has been prepared and issued by Frazis Capital Partners Pty Ltd ABN 16 625 521 986 as a corporate authorised representative (CAR No. 1263393) of Frazis Capital Management Pty Ltd ABN 91 638 965 910 AFSL 521445. The Frazis Fund is open to wholesale investors only, as defined in the Corporations Act 2001 (Cth). The Company is not authorised to provide financial product advice to retail clients and information provided does not constitute financial product advice to retail clients.

The information provided is for general information purposes only, and does not take into account the personal circumstances or needs of investors. The Company and its directors or employees or associates will use their endeavours to ensure that the information is accurate as at the time of its publication.  Notwithstanding this, the Company excludes any representation or warranty as to the accuracy, reliability, or completeness of the information contained on the company website and published documents.​

The past results of the Company’s investment strategy do not necessarily guarantee the future performance or profitability of any investment strategies devised or suggested by the Company.​

The Company, and its directors or employees or associates, do not guarantee the performance of any financial product or investment decision made in reliance of any material in this document. The Company does not accept any loss or liability which may be suffered by a reader of this document.

You may also like