gfp

Always worth a read https://newsletter.semianalysis.com/p/claude-code-is-the-inflection-point?utm_source=post-email-title&publication_id=6349492&post_id=186989157&utm_campaign=email-post-title&isFreemail=true&r=oqhe5&triedRedirect=true&utm_medium=email

I think the only time I actually "averaged down" on Berkshire was the markets reopening after 9/11. I wouldn't have ever owned very much Berkshire if not for averaging up

It sort of depends on what you consider the 'flag pole' heading into the bear flag. I would put the target at a range of prices between $66k and $56k. I would use the 2022 playbook as a guide until the market shows you something is different. So far, the market has NOT shown anybody that anything is different from the 2022 pattern. It is so identical it is comical. Note that in 2022 the market didn't just bottom and turn up to the races - it was a painful slog that convinced a lot of market participants to sell and move on. Only then did it start to rise, eventually becoming another bull market

A bear flag suggests the price goes straight down. Hence the term flag pole. This is the "straight down" portion of the bear flag. I've been posting about this over and over.

That was my point on highlighting the LITE trade in real time on the "chart patterns / technical analysis" thread. Human behavior is WAY more predictable when a widely observed chart pattern is negated than when it "does what it is supposed to do."

You do have to ask yourself how many people could have just googled "hey google, what the hell is a bear flag?" and saved themselves some time

Just nice to see all Warren's picks killing it while he puts himself out to pasture. CB CVX KO AAPL Even some of Ted's picks are having a moment... The firm just might make it after all

This video is a couple weeks old now (Davos) but I watched it last weekend and think it is pretty great to hear from two of the leading minds in AI on stage at the same time with a competent interviewer. Demis Hassabis and Dario Amodei are both primary sources worth listening to - plus Dario's resting facial expressions are endlessly entertaining. https://www.youtube.com/watch?v=9Zz2KrBDXUo

Bought the dip in Fairfax India. Yesterday it was up on Trump/Modi news and today gave it all back.

No I haven't yet. I had a Claude pro subscription to try out Cowork and used it to do some very token-intensive work on my desktop and it did a very complicated task correctly, exactly as I specified. But I also continually ran up against token limits and had to wait and finish after a few hours. So I asked Claude and Gemini how to get my $220 back and they said "do this" so I went and got my $220 back. Then a week later they come out with Claude for Excel and I wished I hadn't cancelled it. So I will probably pony up the big bucks again. I'm not a huge spreadsheet junkie but I do use Excel often and I'm not very fast at it so this tool would be helpful for me personally.

It still looks exactly the same as 2022. Why not make the price action prove to you something is different? A classic traded asset without any supply response to high prices. Highly levered market participants, margin calls, liquidations, plenty of "technical" traders - if any market at all is suited for watching chart patterns it would be this one.

they no listen

LOL ... I gotta ask, is this a reference to a certain homosexual hook up app?

That is correct, Fairfax exercised its Foran warrants back in 2022

It's probably just my market, but everybody in my town with multifamily wants to sell and there is practically no bid on the other side. Rent growth is still negative to zero despite articles claiming nationwide trends are starting to inflect back towards positive low single digits. I'm sure some markets are tight and have nice weather and competent leadership and are doing A-OK. I'll take airport real estate for multibillionaires given a choice.

It is possible to buy the issuer's own shares within a Bermuda reinsurance company (I know that for sure), and Fairfax has several of those (Brit Re, Allied World Assurance, Ltd, etc) and to count un-cancelled de-facto treasury shares toward regulatory capital. But it is extremely unusual and I wouldn't be happy to see Fairfax doing something like that. Let them retire shares and cancel them the "regular way" and be patient - you never know when lower prices for your own shares are just around the corner. I believe Fairfax holds the TRS position on their own shares at the Holding company level.

I don't know, that's not how I invest. It's a moving target. How much upside do I see on Fairfax? $100? Infinity? It's a function of time. I worry about the downside and usually end up with oversize positions and start to shrink the position as the investment works well (starting with the highest cost shares) and then keep my lowest basis shares as long as I can to defer the tax on the lowest basis shares. Only when something screams to me loud enough, like Biglari for a few weeks there, do I rip off the bandaid and pay the tax on those low basis shares. With something like Berkshire or Fairfax I may never sell the lowest basis shares.

From Gemini (which is an AI for you fossils): As of late January 2026, the estimated average electricity cost to mine a single, whole Bitcoin is approximately $59,450. When including additional operational overhead (hardware depreciation, cooling, and facility maintenance), the "all-in" net production expenditure is currently estimated at $74,300. Current Economic Context (January 2026) The mining sector is currently facing a period of significant "margin compression." With Bitcoin trading around $83,000–$84,000 this week, the profitability ratio has hit a 14-month low. Several factors are driving these costs: Post-Halving Dynamics: The 2024 halving reduced the block subsidy to 3.125 BTC, effectively doubling the energy required per coin compared to the previous epoch. Network Difficulty: The global hashrate remains near record highs (approaching one zettahash), forcing miners to use more power to secure the same probability of winning a block. Regional Variance: While the global average sits near $60k, regional costs vary wildly. In the United States, high commercial rates (~$0.14/kWh) have pushed the local cost to mine one BTC as high as $130,000, making many U.S. grid-tied operations currently unprofitable. Cost Breakdown by Efficiency Profitability in 2026 is almost entirely dictated by your electricity rate and hardware efficiency (measured in Joules per Terahash, or J/TH). Scenario Electricity Rate Estimated Energy Cost per BTC Highly Efficient $0.047 / kWh ~$20,000 Global Average ~$0.07 / kWh $59,450 U.S. Grid Average $0.141 / kWh ~$130,000 Export to Sheets Market Implications The Cambridge Bitcoin Electricity Consumption Index (CBECI) recently noted that for 15 of the world's top 20 mining nations, it now costs more in energy to mine a Bitcoin than to purchase it on the open market. This has led to a "miner exodus," where some publicly traded firms (like Bitfarms) are reportedly winding down mining operations to pivot their power capacity toward AI data centers, which currently offer more stable returns on power.

Those 6% subordinated bonds, effectively forced to be called at 5 years, were massively oversubscribed in a very weak week for corporate bond issuance and they upsized the offering from $100m to $150m after receiving $450m in bids. As you probably well know, SKYH borrows almost entirely in the tax free muni market. The JPM warehouse facility, at $200m (up sizable to $300m in the future 'subject to credit approval') was floating -to- fixed swapped to 4.73% for 5 years (also tax exempt to JPM despite being a loan and not a bond offering - very innovative, flexible and low cost capital.) The initial equity requirement for draws on this facility was satisfied with the contribution of a wholly owned campus in Camarillo (CMA) that Sky had purchased for cash earlier. They later added Bradley to the equity component of the borrowing base. You can tell they are doing everything in their power to minimize equity issuance at these prices which they see as a very high cost of capital despite the market not 'getting it' so far. SKYH's CFO came from JPM and was the expert in these type of Private Activity Bonds so I'm sure he still has a contact or two over there. They have also done a few more expensive, small financings in recent weeks with Yorkville, which had tiny equity "vig" kickers for the lender - but this type of borrowing is extremely flexible unlike the bond offerings which have to be used on specific assets within the specific group underlying the bonds. They inflected to cash-flow run rate break even and going forward they should be able to recycle some internally generated equity to meet future equity funding requirements. They would like to see the shares trade above $11.50 a year from now as it would raise the final equity capital they will likely ever need to raise from the markets. Much later, they can drop stabilized campuses down into a REIT if they choose. The hardest part of the entire funnel is getting the ground leases. The rest could be theoretically farmed out, although they won't do that. They are very difficult to compete with at this point. Some of these signed leases took 7+ years of work to get and there is no developable land left for a competitor. They try to get the longest ground leases they can get (with options) but they vary with each negotiation. The fist renewal of the hangar leases after the initial lease term has been averaging very attractive increases in rent and everything steps up with annual escalators from there. NNN leases. Fuel revenue for their home based clients as a kicker. And opportunities for a few minor add on services. You gotta love when your clients are at the very very tippy top of the K-shaped economic situation and if only one or two happen to notice that you've got a hell of a business niche they are certainly sophisticated enough to look into the investment thesis. The fact that EVERYONE is bearish and it's a laughing stock on "Value Investors Club" is icing on the cake.

We'll know something is up when Apple quietly acquires Moscot. Meta already has that relationship with Luxxotica and Google made an investment in Warby Parker. Maybe my hometown Krewe eyewear catches a bid..

I can't' imagine what you would use other than chart analysis of a pure tradable commodity with no supply response or fundamentals. "technical analysis" will tell you what is happening and when what is happening has changed. So far, my chart above is spot on. We will know when it isn't by looking at the damn chart. I'm not going to pull a balance sheet and analyze what's left for owners in a liquidation

Honestly I feel like my growing position in Sky Harbour is sounding fairly contrarian lately given all the hate I have heard on it (which I love!) and a complete unwillingness for even "real estate investment people" to look at the economics of their projects. Cost of capital questions and equity dilution risks have just been largely answered. Quotes from their all-star CEO Francisco in the recent press release like this: They have been masters of raising hundreds of millions of dollars to develop their pipeline without issuing material equity. Subordinated bonds at 6% replace equity in the deals and skyrocket the return to equity on each project. But yeah - totally hated because [SPAC, BOC people, ground leases aren't fee-simple, Airport infrastructure companies only change hands at 6% cap rates because Airport real estate is more rare than beachfront property, SPAC, SPAC, SPAC..]

I found this introduction to Claude for Excel useful. Claude / Anthropic has made deals for a lot of very handy proprietary data feeds (LSEG, S&P Capital IQ, Daloopa/SEC Edgar, earnings call transcripts, Moody's, Morningstar, Pitchbook, ThirdBridge for expert network interviews, etc) that are particular useful to people in our line of work. Made available to Claude pro subscribers on Friday according to this guy https://www.youtube.com/watch?v=f-v0fJgBqhk (youtube link to a video where a dude in a Seahawks jersey explains a bit about the newly widely available Claude for Excel) If you search around youtube there are several intros to this including this guy's: https://www.youtube.com/watch?v=m5YF89Kym1Y (another youtube link) ... Plus some pre-built "agentic skills" that do Jr. Analyst type tasks out of the box: Gemini suggested I might want to use these data feed connectors that Anthropic has contracted for in ways like these:

I'm just basing it on the pattern from 2022. Maybe there should be some adjustment for log scale that means start looking at it at $66k and add as it declines - nothing is every going to be an exact number that is "correct" - but until this pattern stops looking exactly like 2022 I'm not going to assume it will go any differently. When the chart says something different, change your expectations. But for now, the price is trading exactly like it is breaking down from a bear flag in the exact same way as it did in 2022. People like to dunk on charts as bullshit and I'm sure there is plenty of that out there. But people also like to lie to themselves that something is in an uptrend when it clearly isn't and the chart isn't going to lie. This is a downtrend until something changes.

Right - but he is also deeply knowledgeable and working in the industry, publishing his thoughts for free and has been basically correct on everything I have seen him take a stance on previously. I'm not going to trade options on this stuff or whatever he is personally doing with leverage and trading (his tax bill for last years trading is going to be a nightmare) - but I am certainly going to pay attention, especially to his recommendations of who the losers are (which are also going to be popular 'funding shorts' subject to short squeezes like GFS recently had). I like a non - wishy-washy opinion because I can determine if they are correct or not. People who hedge their bets with every statement and never put their capital on the line aren't super useful. Reading his blog plus SemiAnalysis gives decent ongoing coverage of the industry. And Irrational Analysis is more entertaining https://newsletter.semianalysis.com/p/apple-tsmc-the-partnership-that-built?utm_source=post-email-title&publication_id=6349492&post_id=183808784&utm_campaign=email-post-title&isFreemail=true&r=oqhe5&triedRedirect=true&utm_medium=email