August 2024 Investment Update

Dear investors and well-wishers,

Our fund advanced 1.1% in July and with a slightly positive August, is up over 80% calendar year-to-date.

In mid July a broad sell-off hit our sectors, which crescendoed into a something of a flash crash in the first Monday of August, with the third highest VIX spike in data going back to 1990.

Fortunately we had taken profits across semiconductors, which bore the brunt of the unwind.

Even at the time, this seemed like an over-reaction, with the only comparable spikes at the peak of COVID and the financial crisis of 2008. There was nothing quite so consequential this time around.

But you never know.

The headlines focused on rising Japanese interest rates, which triggered a Japanese rush to sell USD investments and bring capital back onshore. But there were three other issues that weekend that made for such difficult Monday trading:

1. A weak unemployment report in the United States

2. Imminent Iranian retaliation after an assassination by the Israelis on their territory (which so far hasn’t come), and

3. A surprise lead by Kamala Harris in the polls (since erased).

After a few negative weeks, this onslaught triggered a buyers strike and a wave of systematic selling.

Coming into this, we knew that algorithmic capital was max long equities, and at some point would be selling aggressively, with the timing and trigger unknown.

There are two species of algo traders that often drive markets (most shares don’t change hands in these events): volatility targeters, or ‘risk parity’, which reduce equity exposure as volatility rises, with Ray Dalio’s Bridgewater being the most prominent, and trend-following algos, which trade short, medium and long term trends.

It’s straightforward to track how these groups are positioned as they are all using the same data. And more interestingly, it’s possible to predict from a given move how much buying or selling they will have to do.

In this case, Goldman Sachs estimated vol-targeters cut exposure from 110% long to 50%, and as short, mid term and long term trend lines broke, the trend followers were sellers too.

Now, with volatility back down, and indices breaking back above major trendlines, these same algos have had to buy back that same exposure.

The last time this happened was November last year, after which we bounced >100%, so hopefully the same thing happens again! Perhaps too much to hope for…

The good news is our risk models gave similar signals, and the rebound triggered re-entry signals across our semiconductor exposure, and it is once again our largest sector weighting, expressed both through the index with a high weighting towards Nvidia and select others.

This is something of a relief, as the fundamental demand remains extremely strong and demand for compute is one of our highest conviction plays, so it’s good to be back. Nvidia reports next week, which will no doubt move the sector one way or another.

Was the employment report that bad?

When employment is rising or falling, it tends to continue. This sounds trivial, but you can’t say the same about most charts in finance, certainly not asset prices: a price moving up or down in a given month gives little indication over what will happen next, and often enough a large movement in one direction is reversed rather than continued.

For employment, this makes sense - if people find jobs in a particular month, they increase spending, and at the margin this leads to more hiring in the next month. And similarly in reverse.

The financial media centered on the ‘Sahm Rule’, which captures this dynamic, and notes that a recession often starts around major trend changes.

The impact on equities is not clear. As Druckenmiller famously pointed out, the best scenario for equities is a weak economy and a central bank trying to stimulate - certainly a possible outcome.

Downwards revisions in job reports suggests the United States is in a weaker position than previously thought. And around the world rate cuts are now priced in.

So recession fears need to be balanced with the fact we are on the brink of global easing cycle.

Which is why we’re going to play close attention to our risk models, and if we have to we’ll cut aggressively. But we’re not there yet.

Stocks

We were also helped by strong results from MercadoLibre, Nubank, Transmedics, and a helpful update from Clarity.

Yesterday Clarity’s imaging drug received Fast Track Designation by the FDA, which is usually for novel drugs addressing significant unmet medical needs.

On the surface that is not the case here, as there are multiple competing radiopharmaceutical imaging solutions already in the market, including Telix’s Illucix. So it seems the FDA sees the benefit of longer half-life, improved specificity, and the ability to detect smaller tumours observed in Clarity’s early trials.

This is our largest position now so will have an outsize impact on our near term performance. We are expecting an update on their does escalation trial this quarter.

Transmedics reported 18% revenue growth on the quarter, +116% year-on-year, and its first cash flow positive quarter. After nine months, this is close to our 2.5x profit target.

Nubank reported customer growth of 60% year-on-year, and 88% growth in gross profit. With industry leading return-on-equity of 28%, after-tax net profit increased 36% on the quarter, and 131% year-on-year. We are up ~3x on the position from our first purchase, targeting a 4x return.

MercadoLibre also had strong results, with 40% growth at an over 10% after-tax profit margin, which is exceptional for its maturity. The company just turned 25.

We took took two new positions in the sell-off.

RxSight makes implantable lenses for cataracts. These can be adjusted after cataract surgery with UV light to finetune the outcome. A more expensive option, but an increasingly popular one. The next generation of their product, giving a wider range of clear vision, is proving popular.

Revenue grew 68% year-on-year, with gross profit growing 102%. Penetration remains low, and with a superior product, the company has plenty of room to grow.

We also participated in capital raisings for Droneshield, Curvebeam and Syntara, which are all now significant positions.

Outlook

In the sharp crash we were determined to close any position which hit our risk levels, and given the overhand of weakening growth around the world, and all kinds of political risks, we will stick to this religiously.

My thesis is the same. We are in a regime that favours semiconductors and select high-growth companies, but the we have moved away from the regime where everything moves up or down together. For the decade leading up to 2022 there were consistent waves which seemed to push all stocks up and down together, with the weaker, junkier companies selling off the most in crashes, but rallying the most in the bull markets.

That’s a sharp difference to today, where the vast bulk of companies are underperforming, and it’s only companies posting results like those above that are moving up. The good news is that strong companies are being rewarded, which isn’t always the case.

Michael