My SIPP has had some good winners recently, with BAE Systems and Rolls-Royce up 24% and 18% respectively in 2026. And Axon Enterprise soared over 20% this week, while Taiwan Semiconductor Manufacturing has doubled in a year.
Unfortunately though, there’s one growth stock that has been bathed in a sea of red for months now. And adding injury to insult, it crashed another 20% yesterday (27 February).
Needless to say, it’s testing my patience. So, should I banish this utter stinker from my portfolio?
Green owl disaster
I’m talking about Duolingo (NASDAQ:DUOL), the digital education app with 85% share of the online language learning market.
Last year, the firm ran a viral social media campaign centred around the supposed death of its green owl mascot, Duo. Perhaps it might do a new one about its own stock, which has crashed 75% inside 12 months. ‘Ouch’ doesn’t do it justice.
AI has weighed on sentiment, as the rapidly improving technology could one day enable a start-up to ‘vibe-code’ a new Duolingo/Spotify/Rightmove/Auto Trader (insert popular consumer platform).
But the firm certainly hasn’t helped itself, with CEO Luis von Ahn sending a poorly worded internal memo last year saying Duolingo would become an AI-first company. This caused a backlash, with some users unhappy with the suggestion that AI would replace humans for translation and content generation.
It’s a case study in how leaning too hard into AI can alienate a loyal fanbase.
On Thursday (26 February), management was alienating shareholders with shocking guidance. For 2026, it expects much lower revenue, bookings, profits, and margins.
von Ahn said: “In 2026, we are deliberately prioritising user growth and teaching better. We’ll focus on improving the free learner experience to grow word of mouth and feed our next user growth engines like chess, math and music, even though that moderates near-term financial growth.”
Moderate? You can say that again. In 2024, bookings growth was 40%. Now, management’s guiding for about 11% growth in 2026.
Glimmers
Duolingo admitted that extra friction it introduced — more ads and upsells — to drive subscriptions has led to a deceleration in daily active users (DAUs). So it will sacrifice short-term profits to improve the free user experience to reaccelerate DAU growth.
There were glimmers of hope. The new chess course has 7m DAUs inside a year. And Duolingo expects its strategy to pay off by reaching 100m DAUs by 2028. If achieved, it would result in a far larger company.
Management expects 20% DAU growth in 2026. Part of this will involve giving AI-driven speaking tools to free users and mid-tier subscribers. It can now do this because AI inference costs have fallen tenfold, and keep falling.
Finally, a $400m share buyback programme was announced, as well as higher-margin targeted ads, potentially in the foreign language someone is learning.
Pull the plug?
Fortunately Duolingo wasn’t one of my largest holdings before this week (and is much smaller now). But as the saying attributed to British economist John Maynard Keynes goes: “When the facts change, I change my mind.”
I got this one wrong.
DIY pensions like SIPPs allow enough time for a potential recovery, so I’m not selling. But due to a loss of faith in management’s ability to create shareholder value, I won’t buy anymore.
