vinod1

I think and I read this somewhere that while AI generated code is probabilistic, the generated code can still produce deterministic output. Trivial example: Use AI to generate code a calculator. The calculator is going to be deterministic in its function. Vinod

It is wrong to think "how much are consumers going to pay?" Consumers are not going to pay much if any at all. It would likely be free to consumers. Enterprises are going to pay. Just like they did mainframes. Because, AI can help companies in a million different ways. We are way past the point of even asking this question. If you look at enterprise processes, it is riddled with manual drudgery, boring processes that can be automated. Health care billing itself is $100s of billion of expense that can be automated. That is one part of one industry in one country. You do not need any new advances in AI. Applying existing AI tech would get you there. The question really is who are the beneficiaries? Vinod

I will state my bias: very pro-Israel. I have been since I was 14 and that is for 4 decades. One of the dozen or so nations that I admire. US can easily win a war on any European country that it wants and occupy them, I assume you are completely fine with US doing that, because you know the European left wing is..... and to the victor belongs the spoils. Also it is the ancestral land of many Americans. That is the argument you are making with the above examples. Yes, these things happened in the past. We do not do these things now. Just as it is wrong to deny Israel the right to exist peacefully, the same rule applies to Israel.

The point I am trying to make is both sides have valid points on many issues. If you start with 7 October 2023, there is no moral ambiguity. Israel has every right to defend with every means necessary and do whatever it takes to make sure it does not happen. Israel is also unambiguously in the wrong with settlement expansion in the West Bank and East Jerusalem. But most issues at least as relates internally to US, the truth is somewhere between what the left and the right has been arguing.

This thread is like arguing over the decimals that come up when you divide 4/11 = 0.36363636363636... Left wing says "there is a 3 after the decimal" Right wing says "there is a 6 after the decimal" Left wing says "that is complete nonsense everyone knows that is a 3 after the decimal" Right wing says "that is what the MSM wants you to believe, there is a 6 after the decimal" Left wing says "You believe whatever Trump says, there is a 3 after the decimal" Right wing says "There is no point in arguing with someone who does not think there is a 6 after the decimal" Left wing says "Good. If you cannot see there is a 3 after the decimal I am glad to stop"

I have a dream that one day companies will not be valued by the yardstick of earnings, nor by the rivers of cash they return to shareholders, but by the share of the companies market cap to the US GDP...

I am including the hedging and short positions as part of the equity portfolio which to me is where they belong. With these the expected return from their equities is essentially pure alpha which as I calculated is not going to get Fairfax acceptable returns. My main point being, alpha did not play as much a role in their performance as their business model. Vinod

Quote Buffett Indicator and Ignore Every Single One of Buffett's Teachings!!!! You know, things like, ignore macro, don't predict markets, focus on the business....

Absolutely. It is unbelievable how much the insurance operations transformed. Before earnings releases way back in 2010, I would be praying for 102% CR. Even Prem used to brag about how the 102% CR was borrowing cheaper than the Canadian Govt! There is not even a glimmer of hope that underwriting would ever be profitable. It is all on investment results. Look at it now. I have to pinch myself to make sure it is not a dream. Vinod

I was talking more about Fairfax portfolio returns in the above post. My post above is trying to assess their portfolio performance relative to indexing. I first got into Fairfax (#1 position) sometime in 2005 and exited in 2011 ($418) Nov/Dec timeframe and bought again in size in 2022 (#1 position). In 2011, it had a snowball chance in hell of returning 15% with the assets/structure they had, now it is the other way around. Vinod

I did an analysis of Fairfax performance in 2015. I analyzed their results from 1986 to year end 2014. It is not as impressive as they make it sound. What the chart Parsad's slide above does not show is impact of portfolio allocation. Say if you have $100 portfolio and invested $10 of portfolio in stocks and they returned 15% but you reduced the stocks from $30 of total portfolio, the above chart fails to capture that impact. That is a very significant drag. Second, their bond performance is not really bond performance. It includes a whole lot of non stock non bond like securities. I compared their performance with what it would have been if they just indexed their portfolio. They added a whole lot less value than most people seem to attribute to them. What is really powerful is their business model. The weak underwriting used to be like a huge hole in their ship, but with that part fixed, just the way they structured the business would lead to good to great returns with just indexing of their portfolio. Vinod Fairfax Returns.pdf

I am thinking of AI evolution as one of two paths 1. Same as any new technology. The technology itself might be disruptive and many companies go bankrupt while many new companies go on to become huge. Agriculture, combustion engines, electricity, cars, trains, automobiles, Internet, mobile phones, etc. Society goes through a period of churn but new jobs and industries develop. Everyone adapts. 2. AGI. Large scale destruction on a scale like the Ice age. Say a company is close to AGI or something similar and it is located in one country. That country would now be able to hack and render every nuclear weapon system of other countries or in someway dominate the whole world in a matter of days. If we take India and Pakistan for simplicity, if one thinks the other is within a couple of days of such capability, it makes a lot of "sense" to use your nuclear weapons when you have a chance to keep the status quo. AI pundits are mildly helpful to read if we are on the #1 path above. If we are on #2 path, predictions are utterly useless regardless of the status of the AI pundit as the path would be so dependent on behavior of millions of people and it is utterly unpredictable. So invest as if we are in path #1, if we are in path #2, your investments are not going to matter anyway. Vinod

During the height of Covid, a family member who is a doctor specializing in infectious diseases stayed at our house for a couple of days when they got a break. The family member was furiously texting at 6 AM to colleagues who are covering patients in her absence. I commented that it must be tough to cover for her absence and the remaining doctors must be overloaded. She told me that they are chatting about SAAS stock tips! Vinod

11 years back I took a detailed look at the historical record of Fairfax investment returns from 1986 to 2014. I was trying to answer a few related questions: (1) What if Fairfax has just invested its stocks and bonds in an index fund right from the beginning, what would its returns have been? (2) What is the alpha Fairfax investment team is adding? How consistent is it? (3) And what can we infer about future returns, especially since bond market returns are going to be terrible due to low yields? I wrote as a blog post at that time in early 2015 estimating Fairfax future returns. This is old and fairly boring, but there might be someone who may find it interesting. The blog (and all 3 blog posts ) has been lost when I transitioned to a new hosting platform. Vinod At what rate can Fairfax compound.pdf

@Viking Thanks for pointing out the changes. You are probably correct and I need to update my views - just like your detailed analysis made me change my mind regarding earnings power. My portfolio returns over the last 4 years owe a lot to you. Thank you! Vinod

Prem has mentioned in the past that they expect 3 out of 5 positions to work out. They were and continue to be probabilistic investors. When you are playing 60% odds, there are going to be periods when you are going to suck followed by periods of outstanding performance. Are we in just such an outperforming period or like Viking says they have "fixed" their approach? What exactly was "wrong" about their investment process before that got fixed? Throwing in the towel a decade after the investment (Blackberry), does not exactly look like they changed or "fixed" their approach. It is just a natural conclusion. I am more of the belief that we are in a period where macro conditions helping Fairfax investments. I am sure they improved their investment process as any good investor would do, but do not see anything that is radically different in their common stock/bond/preferred investments. One area they did change was when buying insurance companies, they did move to quality. I do not see a similar change in their common stock investments. Vinod

A minor change. Appointments: Andy Barnard - Chairman - Fairfax Insurance Group (was President) Brian Young - President - Fairfax Insurance Group (was CEO Odyssey) Carl Overy - CEO – Odyssey Jennifer Allen - Chief Business Officer (was CFO) Amy Sherk – CFO Christine Magee - Fairfax Board Mr. Amitabh Kant - Senior Advisor Viking - Head of Investor Relations

I agree with Castanza and Milu. Why is S&P 500 so hard to beat? I think the most important reason is that it tends to hold on to winners. There is a nice real world experiment, although it is a sample size of 1. There is a fund, Lexington Corporate Leaders Fund is now known as the Voya Corporate Leaders Trust Fund (LEXCX). It was established in 1935, it bought an equal number of shares in 30 major U.S. companies in 1935 and has never bought a new stock since. It only sells if a company goes bankrupt or suspends dividends, and it holds onto spin-offs and mergers. Since inception in 1935, its performance was 10.5% annually, versus S&P 500 11.3% (S&P performance is approximated since it did not exist in 1935). Lexington Corporate Leaders Fund actually outperformed until 2015. Only with the recent 10 year tech outperformance did it lose out to S&P 500. Imagine buying 30 leading companies pretty much willy nilly and matching the market performance and beating vast majority of fund managers over 9 decades? Vinod

You can predict with near 100% accuracy what the position of every single article on WSJ is going to be. Does it increase profits of corporations? If yes, then WSJ is for it. Everything else is just boilerplate around that idea. I read WSJ too and I am as capitalist as they come, but you can spot bias in WSJ very easily.

In India, any business that is connected with a large real estate footprint has risks. To give an example, when Andhra Pradesh was split into two states a decade back, a new capital was announced for Andhra and the ruling party at that time choose a capital and spent billions of dollars on its development. Of course all the land surrounding that capital was bought by people close to the ruling party at that time prior to the announcement. In the next election another party was elected, and they duly chose another city as capital. Again the people close to this party all bought up land prior to the announcement. It is an easy no brainer (heads I make millions, tails I make billions). This is a large scale example, but happens at every level, before a new development area that gets permit, before a company announces a location for its major investment, etc. ("In the election this year there is another change of party and the capital reverted back to the original city." If you make a movie, no one would believe the plot.) At a micro level, a relative bought land a couple of decades back for $20k and it now worth $500k. So the local strongman, just occupied it and wanted $80k to vacate. His argument "You made so much money, why are you being so stingy when all I am asking is $80k and you still make $400k?" He does not even bother to argue that he has some connection to this land or that my relative does not own that land. Going to court means spending the next 10-20 years visiting court once a month. Sure you get justice, it would just take 10-20 years of your life and a few thousand hours of court appearances. To us in North America this might sound too ridiculous, but this is common there. Real estate is an area people understand and a lot of money is made from development of previously undeveloped areas. You need to be politically connected or have muscle of your own. Or you might find that your real estate suddenly has problems. I have no idea of Fairfax's connections and if this even a relevant risk. Just the idea of a 2nd airport, introduces long term risks. As long as BIAL does good, but not spectacularly good, this might not be a risk. A 2-3x over several years might escape attention, a 10x-20x would. I am a cautious pessimist ("someone who anticipates negative outcomes but I am open to the possibility that it all works out"), and also uninformed about all the politics around this. So this might be utterly and completely ignorable risk.

You most certainly do! If you say something like "I would pay 10x or 15x owners earnings or FCF", you are doing DCF. You have just buried it underneath a whole bunch of assumptions. What you really mean is you do not write down Year 1, year 2, year 3 cash flows and perpetual growth, etc. But when you put a multiple on owners earnings, you have a implicity assumed all of these. Vinod

The argument is they can build a ladder of 1 year to 5/6 year ladder that roughly matches liabilities. That is it. If interest rates are say, below 2%, it is fine to say, no that is too low and I am not taking any risk. You are just balancing reinvestment risk with this approach. Vinod

TIPS are widely traded and available in size. There is no real market constraint if Fairfax wants to hold a portfolio of inflation-linked bonds that match liability duration. The fixed income earnings would then fluctuate with inflation rates, compared to having a very predictable 2-3 year fixed income earnings now. Vinod

I am sure you are aware, but there is something called "Disruptive Innovation" vs "Sustaining Innovation". If there is disruptive innovation, usually the incumbent companies do not want to cannibalize their profits and it leads to innovators overturning the incumbents. If innovation is sustaining, incumbent companies go on to dominate. Internet ("Dot Com") was definitely disruptive. Many think, and I agree, that AI is more sustaining innovation and existing incumbants are going to be dominant. US tech companies going all out on AI is much more preferable to resting on their laurels, buying back 3% of their stock every year and calling it a day. If google spends $100 billion for next 4 years that is less than 4 years profits. So pay 4 PE less if you think that is all 100% wasted. If they win and win big, say hello $10 trillion. Vinod

Those are the wrong people to talk to! The core engineering software work at Mag7 and the like is vastly different from the business applications that make up 95% of the software work in the economy. The vast majority of the software development work is not like the work at Mag 7. Talk to the average software engineer at a financial or non-software company, and you get a completely different picture. Vinod