Curious about everything, especially markets. I share my thinking to find out what I’m missing. Thoughtful pushback welcome. Not financial advice.
Joined October 2017
- Tweets4K
- Following624
- Followers1.6K
- Likes11.2K
Pinned Tweet
Below are my thoughts on why $LMND continues to get crushed even after what I consider a good 3Q26, where the metrics that matter most to the business model long-term moved in the right direction.
The only fly in the ointment in 3Q26 were the LR comments on pet and Europe, but $LMND made rate filings months ago to address these. This is the kind of temporary fluctuation that all insurers deal with in the ordinary course of business. Nothing thesis changing here.
And yet the stock trades like death. It is always possible I am missing something, but my working theory is that the company put itself in the penalty box earlier this year by overpromising, blowing up their own guidance and their supporters in the process.
$LMND started the year guiding for flat opex, even with the structural upward pressures of other insurance expenses that are proportional to revenues. But then they also saw an opportunity to make strategic hires to take advantage of the progress in AI to significantly accelerate their geographic and product rollouts. Between those two things we got a huge spike in expenses that blew up the prior guidance.
Ironically, I believe they made the right business decisions. We want them to invest to maximize long-term business value. The ongoing accelerating in growth with improving marketing efficiency suggests so. The problem was not the execution. The problem was setting unrealistic expectations that they would have failed to meet even without the extra investment.
Put differently, if (even economically rational) decisions violate prior expectations set by the company, it makes investors ask why the company missed their self-imposed targets and whether the underlying model is broken. This happened in two consecutive quarters earlier this year. That is why $LMND is in the penalty box.
Companies don't need to tell every bullish story under the sun. Since $LMND is a fast evolving business (especially with the accelerating progress in AI), they need some room in the guidance to make the most value accretive business decisions day to day, even if those sometimes hurt short term numbers.
The answer is not to guide on things that you cannot be reasonably certain you can deliver with the normal volatility in the business and investment opportunity set. Instead, be accountable for the top and bottom line guidance you do give, and deliver on it with no excuses. No drama beat and raise stories are good. Long explanations after the fact are bad. Guidance on other metrics is good only as long as you are willing to be accountable for them under a no excuses policy, even while the world around you evolves.
The business model of a structurally low cost provider of insurance was fabulous even pre AI (just look at the track record of value creation at Geico and Progressive). And $LMND is super charging this with AI in a way no one else is. You do not have to promise any more than that.
The Investor Day will be important. It is a real opportunity to reset the narrative. I hope that the company sets reasonable, first principles, achievable targets, and then just gets to work on beating, raising, and rebuilding investor confidence.
I am bullish on Nick Stead, the incoming CFO. Even though he is a young guy just stepping into the role for the first time, his comments on the last few earnings calls suggests that he gets it.
I am also glad that @daschreiber acknowledged the other insurance cost pressure in his interview with @PaperBagInvest earlier this year. It suggests that the company understands they should not promise things that are unsupported by the fundamental economics of the business.
The reaction to 3Q26 earnings suggests that investors have PTSD over this stock. And anyone who as consistently defended $LMND looks like a bit of an idiot right now, given where the stock is. That is why the stock is having a hard time finding a bottom, imo.
Personally, I don't often fight the tape like this, but I currently have my largest long position in $LMND ever. I think that $LMND's business model is special and has the opportunity to deliver tremendous long-term value. And I think that within a couple of quarters the noise from earlier this year will fade as opex growth returns to a normal trend rate. And maybe growth continues to accelerate as a result of this year's investments as well.
I have been early, and it has been painful, so feel free to take my opinion with a grain of salt. But I also think it is our job as investors to stand up for good companies going through hard times.
I look forward to the Investor Day and to brighter days ahead.
cc: @shai_wininger
Andreas Schulz retweeted
Goldman S&T: "Yesterday was one of our busiest days of the year on the equities trading desk (from a number of trades and notional executed perspective). Today is the exact opposite. Our phones are exploding with asks for color. Unfortunately we aren’t seeing much. The long only community is frozen (again). There was massive desrisking into midterms and traders were surprised by yesterday’s move and got sucked back in. HFs selling back pockets of the market (hyperscalers) that they bought yesterday."
Andreas Schulz retweeted
I have covered hedge funds for nearly 16 years, and I can't remember a day like today. Here's what happened in the space of about ten hours:
Good post summarizing how we got to a day like today...
xitter.cf/huaxying/status/211842…
The HF game used to be simple. Find the nugget the sell-side mismodeled, put on the trade, wait for the revisions and the long-onlys. You knew how being right fundametnally got you paid.
Then came alt data, and the game moved up a level. You didn’t have to wait for the company to report. You could watch spending, shipments and foot traffic to see the quarter taking shape. For a while, having the dataset first was a very good business.
But over time, the advantage became harder to hold onto. More data, better tools, more people digging into smaller pieces of the market. Funds understood companies better than ever. And somehow it got harder to make money from that.
I think that comes back to a first principle of making money in the market: someone has to follow your trade. And if you look around the table, the players have changed:
- ETFs grew. More money buying themes, styles and characteristics rather than company fundamentals. In cap-weighted funds, yesterday’s winners became a bigger part of what the next dollar bought.
- Pods multiplied. More teams with similar data, tools and risk constraints, getting better at finding the same nugget and putting on the same trade.
- Long-onlys rotated. Capital rotated from value into growth, creating a split: in sectors losing capital, even strong quarterly results could struggle to bring those buyers back. In sectors getting the flows, being fundamentally right came with an extra tailwind.
- Retail grew. Some through ETFs, some through individual names, bringing another pool of money responding to stories and price action rather than the line item you had figured out.
I saw it firsthand in financials in the mid-2010s. Energy had a similar problem. Long-only capital was leaving. Pods were still doing the work. You could catch the revision and still be waiting for a buyer who had left. Those names became “hedge fund hotels”: everyone knew the company and management well, had similar models, and obsessed over the same one or two line items where they thought the alpha was. Everyone was playing the same quarterly catalyst, but often, even when you were fundamentally right, there just wasn’t anyone who cared enough to follow.
Then the same dynamic hit sectors with a lot of alt data. The apparent alpha attracted more people, while providers increasingly sold both the data and the analysis. Everyone had roughly the same read on the quarter, and it got into positioning earlier. Being early used to mean at least the other HFs would follow your trade. Now they were already in, and the long-onlys weren’t playing the quarterly game. Nailing the print didn’t necessarily mean a big payday. But missing it when everyone was already on the same side could mean a big loss.
Through a lot of that period, tech and healthcare generally worked better. Tech had fundamental winners with growth allocations and ETF flows behind them. Healthcare, especially biotech, had events that could force a reprice. Being right could still pull in new capital.
But the same thing that made tech work so well also made the trade more crowded than ever, especially as the AI trade took hold. Money was coming in from almost every direction: retail, ETFs, long-onlys, trend followers and fundamental hedge funds. This time, the hotel wasn’t just full of HFs waiting on the same quarter. Pretty much everyone had checked in.
You can see the pattern in the table below. Different teams, different sectors, all that company-level work. Yet when momentum moved sharply in the selected months, a lot of the pod shops’ P&L moved with it. (Citadel does seem to hold up better on the downside.)
So why do pods often move first when the trade turns? Their business model runs on leverage. Tight drawdown limits mean they can’t sit through the same losses (-2% is a lot!).
Those earlier sector unwinds were mostly hurting other pods. But this time, the same names sit in far more portfolios. If pod deleveraging pushes prices down enough, it can trigger selling from single-manager funds, quants, (levered) retail and other holders too. How far it spreads depends on each group’s risk model and how much pain it can take.
It’s a different game now. The hard part is spotting the break when, most of the time, following the crowd pays. Can AI help, or does it just make everyone better at finding the same trade? I have a few thoughts, but this post is already long enough, so more next time…
This latest decline in $LMND is quizzical, to be polite. This was a great quarter by any metric that matters.
There have been quarters in the past where I understood a weak market reaction. This is not one of them...
Congrats to $LMND. The quarter looks like a quality beat across the board.
(1) Q/Q IFP Growth accelerated
(2) Marketing efficiency up Q/Q and Y/Y
(3) ADR up to 86%
(4) Opex ex growth spend and synthetic agent interest down Q/Q
Hard to nit pick here...
xitter.cf/shai_wininger/status/2…
12 straight quarters of accelerating growth 🚀 @Lemonade_Inc’s Q3 2026 results are in!
Highlights (Y/Y)
🚀 12th consecutive quarters of accelerating IFP growth
🚀 IFP reached $1.55 Billion (+34%)
🚀 Revenue grew 73% to $336M
🚀 Record Gross Profit of $118M (+48%)
🚀 3.52M customers (+23%)
🚀 $27M adj. Free Cash Flow → 6th consecutive positive quarter
🚀 $1.2 Billion in cash + investments
Car picking up speed
✅ Car IFP grew 63% Y/Y
✅ Now in 14 states, covering more than half the U.S. population
✅ Autonomous Car for Tesla FSD now in 6 states
Lemonade OS
✅ >2.5x the premium base, with fewer employees than four years ago
✅ 25 new product-geography combinations launched this year
✅ Renters now available nationwide!
✅ Native support for personal AI agents coming soon
✅ Live demos of Lemonade OS at Investor Day, Nov 17
Up next
⬆️ Guiding to our first-ever positive adj. EBITDA quarter in Q4!
⬆️ Reiterating our expectation of a full year of positive adj. EBITDA in 2027
⬆️ Raising FY revenue and gross earned premium guidance
Morgan Stanley provides some high quality content through their X account for free. Worth a follow, imho...
xitter.cf/MgnStnlyDerivs/status/…
Yesterday we highlighted that the volatility market was pricing a small potential move in the S&P 500. The index ended up 1.6% on Nov 4th, well beyond the 1.2% move implied by the volatility market.
The Momentum factor* fell more than 6%, placing in the bottom 1% of all daily returns in at least a decade. Prior winners fell almost as much as prior losers rallied. Semiconductors fell below July lows while Consumer Discretionary outperformed Energy by over 9%.
Remarkably, the Beta factor* is down almost 3% despite the sharp rally in equities. This comes on the back of its worst month in 15 years, with Beta falling 22% in October.
YTD stats are striking. Beta is down 11% and Momentum 14%, while the S&P 500 is up 12%.
The simultaneous selloff in Beta and Momentum has echoes of July, but the drivers are different.
July was the worst month for Momentum in 15 years, driven by a positioning unwind as investors sold prior winners. October was the worst month for Beta in 15 years, driven by a broader macro shock as tightening financial conditions punished high-volatility stocks.
* Morgan Stanley Institutional Equity Division is measuring Momentum here as the performance difference between the top 200 outperformers vs. the bottom 200 underperformers of the Russell 3000 over the past 12 months with a 1-month lag. Beta is measured as the performance difference between the top 15% of liquid Russell 3000 members based on 6 and 24-month betas and the bottom 15% on the same metric.
Well done Korea!
xitter.cf/ProducerCities/status/…
"South Korea emerges as the best-performing advanced economy since 2019 if you adjust for population growth, followed by Greece. The US slips to third place" giftarticle.ft.com/giftartic…
Hear, hear...
xitter.cf/A_May_MD/status/211798…
I hope (and think) we are likely to be entering a prime biotech deal-making spree over the next few months.
Pharma valuations have been soaring (and seemingly will continue to do so today), all while quality biotechs that *could* be M&A targets have been relentlessly abused over the last several months.
BTW, side note: If you're looking at the $XBI to determine that "Meh, it hasn't actually been that bad for SMID bios. Look at $XBI in the $150s!!!" - you are absolutely missing the true picture of what has been going on....badly. FORGET the $XBI you think you know - that index is long dead, filled with large cap stocks that DO NOT represent what $XBI used to represent. *And* consider the components that have driven *all* of the upside in the biotech indices of late...it has almost exclusively been GARBAGE that has rallied, leveling out the appearance of damage on the surface, all while under the hood the "quality" stuff has been absolutely demolished.
Let's review the setup for an M&A frenzy here...
➡️Pharma revenues have been looking great. The cash is there, and the confidence in the business is there.
➡️Pharma stock prices are looking even better, despite what has actually been a pretty abysmal series of high-impact clinical trial failures from pharma since around the middle of this year.
➡️The patent cliffs are still there, largely unaddressed. The biggest in history.
➡️Not only that, but the cliffs are actively steepening, because the revenue streams that have (largely) been driving pharma valuations higher and higher are the very ones that are about to run out...making the need to act and fill those impending gaps even more dire (there is MUCH more for a pharma to defend going toward a patent cliff when their stock is +50% YoY - dire loss of revenue streams is simply not being priced in recently...they have to act the make that an actual reality)
➡️The elections are now out of the way and we basically have the guarantee of gridlock for at least 2 more years (healthcare's absolute favorite setup).
➡️Rates/cost of capital are almost certainly coming down as part of last night's election results
➡️The 2 biggest healthcare conferences of the year happen within the next 3 months, and (hopefully) pharma will be eager to keep their momentum going by having comprehensive plans for how they plan to manage their patent cliffs ready to go for investors by JPM...
➡️All the while, "quality" biotech M&A targets have actually been in a legitimate bear market over recent months (again, SPARE ME with references to $XBI - this is a borderline irrelevant datapoint today). This is perhaps the most interesting facet of the setup to me. It seems that absolutely everything points to a near-term resumption in biotech M&A after a several-month drought, yet the M&A targets continue to price in significant skepticism...This *further* incentives deals to happen, as valuations are quite palatable for pharma. Somehow, pharma's desperation can be fully evident WITHOUT the market eroding their bargaining power by bidding up their M&A targets. Really a wonderful setup for deals IMO.
Pharma is flying as high as ever despite many recent clinical-stage setbacks. Gridlock grants them a political tailwind. Their patent cliffs not only draw closer with time, but also steeper with their ever-growing revenues and valuations, which have known expiration dates. All the while, the solution to their problem is no secret: M&A. Yet, the very obvious targets of all these aligning incentives have traded down relentlessly over recent months...Pharma gets the solution to their problems on sale right now despite the fact we all can see they are becoming increasingly mandatory buyers?
Let's hope this is the perfect storm for some serious deal making over the next 3-4 months...the setup certainly looks ripe to me! I predict we will see rapid and *sharp* changes in biotech M&A target valuations as soon as we see a couple decent deals take place...Fingers crossed!
Andreas Schulz retweeted
@Lemonade_Inc Renters is now available nationwide.
Minnesota went live today, completing all 50 states.
(1/3) Another day, another breakthrough in AI. If models can reason about weights rather than discover them from massive pretraining optimization runs, won't we see a much faster takeoff at lower cost?
The implications seem pretty earth shattering...
xitter.cf/rikiyatakehi/status/21…
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math.
rikiyat.github.io/blog/weight-sp…
github.com/RikiyaT/lm_und…
huggingface.co/Qwen/Qwen3.5-0…
huggingface.co/RikiyaT/Qwen3.…
(4/3) One more thought. So far AI was many orders of magnitude less energy efficient than the human brain, which does all it does with about 20W of power.
If AI can now custom write these weights to do only exactly what is needed exactly accurately, will this gap close?
Market structure has changed a lot since the .com bust, and in general the pace at which everything happens keeps accelerating.
Still an amazing chart...
xitter.cf/AndreasSteno/status/21…
Maybe the next leg of the AI trade is not in the infrastructure names, but in the beneficiaries of AI driven productivity?
xitter.cf/GavinSBaker/status/211…
Hard for me to look at these charts from the FT, especially the first one, and not see some evidence of “ROI on AI.”
Seems like a productivity surge post-Covid era overhiring that began to reverse before ChatGPT came out and then resumed its upward trend after ChatGPT.
Especially given the @a16z data on revenue per employee at their portfolio companies and the @EPBResearch data showing that the 3 industries that have most heavily adopted AI have gone from 2.2% productivity growth to 5% productivity growth.
And makes sense that it is showing up in the US first - America is 72% of Anthropic revenue vs. 50%ish at most application software companies.
Great stuff from @jburnmurdoch even if I disagree with his conclusion.
De-grossing is one of the least satisfying explanations of a drawdown. But nonetheless, sometimes it is the explanation...
xitter.cf/MgnStnlyDerivs/status/…
The unwind in Hedge Fund equity exposures this year has been the biggest since at least 2017.
Active de-grossing (reduction in long + short positions) by global hedge funds has now hit over 8% of total gross exposure since June, according to the Morgan Stanley Prime Brokerage Content team.
This has taken gross leverage for US long/short funds down to the 10th percentile versus the last year.
November will be a big month for $LMND, with earnings and investor day. Anything could happen here, and from past experience it will be a big move one way or another.
But longer term I feel good that $LMND will take advantage of the AI takeoff...
xitter.cf/levie/status/211602991…
It’s almost incomprehensible how wide the distance is between someone who’s working at the frontier with AI and basically everyone else right now.
The crazy thing is these lines are not converging, but only growing farther apart. If the curve of AI progress had leveled off, it would certainly have given most people and companies time to catch up or implement their workflows in a standardized way. The technology would have neutralized competitive advantage because everyone would be using it the same way.
But what’s happened is the opposite. The curve of AI progress is basically going vertical right now, which means if you had codified your processes 3-6 months ago you’re likely already behind again. Now imagine most companies that kicked into gear a year ago and thought that they were done. Even further behind now.
I’ve never seen a technology in my life and career that improves at this rate, which consequently means you need to upgrade your expectations at a similar rate. Our brains weren’t wired for this, and our organizational change management processes certainly weren’t either.
With any form of recursive self improvement, things just get wilder. It’s a time when you will need to be constantly reevaluating the things that didn’t work just a couple months ago to see if they now can be solved. AI native companies certainly have a strong advantage due to this right now, so now it comes down to can the bigger companies keep up. Wild times.
