Bombshell from $APX this morning. Full text of the release is below.

This kind of announcement has been foreshadowed by many here, so it will not really be a surprise to them. However, the market has taken a sharp intake of breath with SP down almost 40%. This says that some holders remain in denial that this business model has had its day.

I first took a position in $APX in December 2016 as part of the MF PRO recommendations, like many here. And over the next year built a sizeable position premised on the growth of demand for data to train AI models. I'll admit to being initially skeptical when some started saying the model would be short-lived, but as the chorus of respected voices grew, I progressively exited (missing the peak on the way down) and pocketing between $12.90 and $39,00 over the period Feb-2020 to July-2021. How long ago that now seems.

Today's announceent is surely a very important milestone. Full text follows.

As a personal observation, I teach business operations as one of my occupations. A current student of mine last week showed how he coached ChatGPT 4.0 to solve a rather complex analytical problem I set for the class. The level of coaching was non-trivial, but the progress LLM's have made in one year is remarkable.

ASX Announcement

----------------------------

Appen Limited (Appen) (ASX: APX) received notification on Saturday, 20 January 2024 AEDT from a material customer, Google LLC, that as part of a strategic review processit will be terminating its global inbound services contract with Appen, resulting in the cessation of all projects with Appen by 19 March 2024. Appen had no prior knowledge of Google’s decision to terminate the contract.

In FY23, Appen’s revenue from Google was $82.8m1 at a gross margin2 of 26%.

The news is unexpected and disappointing, particularly considering the progress made against Appen’s transformation and performance in November and December 2023. Appen saw Q4 on Q3 growth in both Global Services and New Markets (including China). On a YoY basis, Global Services Q4 2023 revenue was down while New Markets (including China) was up. Within the New Markets division, China achieved a quarterly revenue record in Q4 2023 of $11.1m. At a group level, based on unaudited management accounts, Appen recorded revenue of $24.1m and $25.9m in November and December 2023 respectively, along with execution of cost management initiatives, which enabled achievement of cash EBITDA breakeven objectives.

Based on the unaudited management accounts, Appen also recorded:

• Underlying EBITDA3 (excluding FX) of $3.2m and Underlying cash EBITDA (excluding FX)4 of $2.3m in December 2023;

• Revenue of $273.0m and an Underlying EBITDA (excluding FX) loss of $20.4m for the full year; and

• $32.1m cash on hand at 31 December 2023.

These results are preliminary and subject to change as the audit process is finalised. Appen’s full year FY23 results remain unaudited and are subject to Board review and approval as well as completion of the external audit.

Appen continues to focus on cost management, business turnaround and delivery of high-quality AI data for its customers. Appen will immediately adjust its strategic priorities following the notification of the Google contract termination and provide further details in its FY23 full year results on 27 February 2024.

Authorised for release by the Board of Appen Limited.

APX - Over and Out

First of all, a big shout out to my fellow members for the various notes and references which I have digested. As a long term investor, I am a reluctant seller, and resisted the desire to sell at the market opening yesterday after having listened to the investor call.

My valuation scenarios are $11.00 base with downside $9.00 and $17.00 upside, with none of the cases assuming a fundamental change to the business model and competitive positioning we are seeing now.

In all cases, I assume that automation of data-labelling grows progressively and over time captures the lion's share of the AI-data opportunity. However, I assume that a signficant and growing niche for manual labelling persists and that APX continues to develop its own limited automation capabilities. 

In terms of revenue growth scenarios, I settled on 10% as a central case: established markets are stable at best and new market growth starts out at 25% from a low base and also matures over the decade, due to competition. (Evidence: yesterday's call cited autonomous driving in China as a new market, but based on the recent Tesla AI day, it will not take long for automation capability to reduce the opportunity here too.)

In essence, I believe APX has a place in using manual labelling enhanced with productivity tools to take on emergent and niche customer opportunities, particularly where those customers cannot access advance data-labelling tech and rely on manual data-labelling at the early stage and for validation. However, the more material of these project will warrant development of automation, which over time will become ubiquitous.

Yesterday's Q&A was not a pretty sight. CEO Mark Barayan seemed almost weary in defending the guidance on 2H21. The pattern of a strong 2H is well-established, and it sounds like they have got close to their customers and are confident in the revenue. But guidance for 2H21 is well beyond what has been achieved in the past. Success gets APX towards 10% growth longer term, which is my starting point.  But it is unclear where the further upside and return to higher growth is coming from. I have no doubt that he has been beaten up a lot in the last 12 months from instos and Board. A final straw for me was the re-issuing of new guidance, and then in the same breath guiding towards the bottom of the new range. Surely, no-one can believe that the risk profile around H2 delivery is weighted so finely that you wouldn't put your target in the middle of the range? I consider there is too much pressure for another "miss" here.

A comment on my $9.00/$11.00/$17.00 scenarios- this doesn't mean the upside is asymmetric and that I should HOLD. These are more sensitivities in the modelling than true scenarios. I have not valued the scenario where the business model is broken (see some of the commentary from other Members) and we see lower growth and lower margins. You can create anything from $5-7 in that case, but I have chosen not to. I also did not seriously consider scenarios whereby APX invests significant capability in automated data labelling. It looks like that game is over and big tech has got their first.

I look forward to returning to this note in 12 months time, as I think we will learn a lot over that period. I am not sure I would have done much differently based on what I know today, but will review the lessons with time.

[DISC: Not Held.]

I exited APX this morning at $10.81. I first bought APX at  $2.75 on 9/12/2016. Enjoyed the ride up and took profits along the way and lost some of the returns on the way back down. Overall, a c.100% return over 4.5 years, which is better than my portfolio average.