Iron Dome missile defense; using GPTs & LLMs in economic research; very briefly noted; not the Efficient Market Hypothesis; & The Neofascist Five Supreme Court Justices: America's Downward...
In the matter of "how your ideology shows", one could characterize the Fama-French program as "research" or, just as defensibly, "fiddling with model parameters".
I do object to describing a factor model as "explaining" variance in a text intended to be read by the general public. This term is perfectly intelligible - taken for granted - by economists, but is likely to mislead the general reader. And perhaps, subconsciously, it misleads the professional economist too. Suppose, arguendo, that you accept the five FF factors as stable, persistent, and fundamental. The thing to *explain* is why. For example, if value stocks really outperform growth stocks, why is that? An example that has the correct form of an explanation is "value is defined as book-to-market ratio, and there is regression to mean". Another is "the cash flows from value stocks are more evenly distributed than those from growth stocks, therefore they have a higher interest rate convexity". Yet another is "value stocks outperform growth stocks in bad states of the world; therefore the stochastic deflator values them more, ceteris paribus." *Research* would attempt to determine whether and to what degree these explanations are correct. Of course, if the effect is emergent and transient after all, that would be research founded on sand, wouldn't it.
Anyway, "in my factor decomposition of the covariance matrix, factor X accounts for Y% of the total variance" does not have the correct form of an explanation.
'‘Nvidia’s stock has been soaring… pick and shovels… betting on the infrastructure that underlies a new technology…. “There seemed to be a railroad bubble… one a decade after the Civil War,” said Brad DeLong, an economic historian at the University of California, Berkeley. 19th-century investors rightly sensed that railroads were about to revolutionize the economy, and still lost their shirts. But just like the telecom companies that laid internet cables in the 1990s, many of which went bankrupt, the failed railroad companies gifted the economy some helpful plumbing. “In the aftermath of each railroad bubble, we were left with an awful lot of railroads that had been built by investors who had gone bankrupt,”'
I am not clear who the target of this analogy is. The railroad companies went bankrupt, but did the steel manufacturers of the rails? Some telecom companies followed, but did the manufacturers of the fiber optic cable, routers, etc. also go bankrupt? The suppliers of picks and shovels, and other supplies did well during the gold mining booms, but as long as they had other businesses, did they go under? Or did they do well by being middlemen, causing stress on the manufacturers of those supplies when the booms ended?
In the tech industry, some of the big failures were the computer manufacturers - DEC, Silicon Graphics, and Sun Micro, to name just 3. IMO, they failed because they had expensive products that could be replaced by cheaper "good enough" competitors. But the chip makers all seem to have done well and remain active. While Warren Buffett famously said he would never invest in companies requiring fabs as they never created enough free cash flow, TSMC has done very well, and ASML that supplies the lithography equipment remains highly profitable. Intel's fortunes have faded for similar reasons to the computer manufacturers. CISCO did well in the internet boom, but despite leaner times after the dotcom collapse is still doing well.
Is Nvidia like TSMC or Intel/Sun? At this time they have no serious competitors for their technology. Arguably their technology could replace a lot of CPU-based systems in data centers. Their graphics cards are often the most expensive part of any desktop/gaming system. So while they are going gangbusters during the AI boom (with a valuation reflecting it), are they doomed to collapse when the boom eventually slides into the "trough of disillusionment" or will demand for their innate parallel processing approach see steady growth of their chips and systems? As computing resources are now a significant energy sink, technologies that reduce power consumption from ARM chip technology to GPUs is a welcome development.
After a bust, what technology remains useful at firesale prices? Railroad lines were torn up or left rusting, replaced with shiny new roads. It would ne nice to think that all that "dark fiber" is now lit up, but is it? Are all those comsat swarms in LEO going to be useful if SpaceX at al fail to make enough money from subscribers, or will they just be allowed to burn up on reentry without replacement reducing the swarm's capabilties?
IOW, is the basic idea too simplistic and reality far more nuanced, especially when technology is rapidly developing?
Levin: What is the model here? People invest in projects that have NPVs evaluated at externality-inclusive costs and benefits > 0 but with only private costs and benefit NPV < 0. A glorious outcome – exactly the opposite of what we normally worry about -- but why does it happen? Would it be better in these cases if the externalities were internalized? Should, could more such mistakes be encouraged? How?
Del Mastro: It is the case that Neoliberals have had to commit to lots of "affordable (with subsidies) housing" to get Progressives (who, just like "conservatives," see everything as zero sum) to go along with land use and building code reform. Just arguing that land owners, builders, city budgets, new residents and new businesses will benefit is not enough!
like many second best policies, the dose makes the poison. A low fixed percentage that does not have to be re-litigated investment by investment would be pretty harmless.
Levitz: Right conclusion but in need of two corrections.
a) We really cannot be sure if when and whose wages are rising faster than inflation. BLS produces unit value indices of wages, not price indices.
b) The concern about wages causing inflation is wrong not only because only the Fed causes inflation but also because it implicitly assumes that the only relative price stickiness that Fed engineered inflation needs to lubricate is goods/wages.
Rodrick: "Central bank independence" is a thing. It is possible that it be good (even a silver bullet) or bad. "Industrial Policy" is not a thing. There are a gazillion arguments for favoring one kind of firm over another, most bad, some good, but It just is not a thing that could be good or bad.
Of course Neoliberalism is a kind of ideology and I'm proud to show it.
You do realize that the failure of liberalism after WWI resulted in the 1920-30s shift towards fascism in Europe and Bolshevism in Russia. However proud you may be to be neoliberal, it won't help if history repeats itself and we end up as a fascist state in the USA. Idk what the solution is, but unless we can convince enough voters that they are not living in a socialist dystopia requiring a "strongman" to save them, then we could be in for a world of hurt (unless you are a white, Christian, heterosexual, male, and preferably with a family ancestry rooted several generations or more in the USA).
In the matter of "how your ideology shows", one could characterize the Fama-French program as "research" or, just as defensibly, "fiddling with model parameters".
I do object to describing a factor model as "explaining" variance in a text intended to be read by the general public. This term is perfectly intelligible - taken for granted - by economists, but is likely to mislead the general reader. And perhaps, subconsciously, it misleads the professional economist too. Suppose, arguendo, that you accept the five FF factors as stable, persistent, and fundamental. The thing to *explain* is why. For example, if value stocks really outperform growth stocks, why is that? An example that has the correct form of an explanation is "value is defined as book-to-market ratio, and there is regression to mean". Another is "the cash flows from value stocks are more evenly distributed than those from growth stocks, therefore they have a higher interest rate convexity". Yet another is "value stocks outperform growth stocks in bad states of the world; therefore the stochastic deflator values them more, ceteris paribus." *Research* would attempt to determine whether and to what degree these explanations are correct. Of course, if the effect is emergent and transient after all, that would be research founded on sand, wouldn't it.
Anyway, "in my factor decomposition of the covariance matrix, factor X accounts for Y% of the total variance" does not have the correct form of an explanation.
'‘Nvidia’s stock has been soaring… pick and shovels… betting on the infrastructure that underlies a new technology…. “There seemed to be a railroad bubble… one a decade after the Civil War,” said Brad DeLong, an economic historian at the University of California, Berkeley. 19th-century investors rightly sensed that railroads were about to revolutionize the economy, and still lost their shirts. But just like the telecom companies that laid internet cables in the 1990s, many of which went bankrupt, the failed railroad companies gifted the economy some helpful plumbing. “In the aftermath of each railroad bubble, we were left with an awful lot of railroads that had been built by investors who had gone bankrupt,”'
I am not clear who the target of this analogy is. The railroad companies went bankrupt, but did the steel manufacturers of the rails? Some telecom companies followed, but did the manufacturers of the fiber optic cable, routers, etc. also go bankrupt? The suppliers of picks and shovels, and other supplies did well during the gold mining booms, but as long as they had other businesses, did they go under? Or did they do well by being middlemen, causing stress on the manufacturers of those supplies when the booms ended?
In the tech industry, some of the big failures were the computer manufacturers - DEC, Silicon Graphics, and Sun Micro, to name just 3. IMO, they failed because they had expensive products that could be replaced by cheaper "good enough" competitors. But the chip makers all seem to have done well and remain active. While Warren Buffett famously said he would never invest in companies requiring fabs as they never created enough free cash flow, TSMC has done very well, and ASML that supplies the lithography equipment remains highly profitable. Intel's fortunes have faded for similar reasons to the computer manufacturers. CISCO did well in the internet boom, but despite leaner times after the dotcom collapse is still doing well.
Is Nvidia like TSMC or Intel/Sun? At this time they have no serious competitors for their technology. Arguably their technology could replace a lot of CPU-based systems in data centers. Their graphics cards are often the most expensive part of any desktop/gaming system. So while they are going gangbusters during the AI boom (with a valuation reflecting it), are they doomed to collapse when the boom eventually slides into the "trough of disillusionment" or will demand for their innate parallel processing approach see steady growth of their chips and systems? As computing resources are now a significant energy sink, technologies that reduce power consumption from ARM chip technology to GPUs is a welcome development.
After a bust, what technology remains useful at firesale prices? Railroad lines were torn up or left rusting, replaced with shiny new roads. It would ne nice to think that all that "dark fiber" is now lit up, but is it? Are all those comsat swarms in LEO going to be useful if SpaceX at al fail to make enough money from subscribers, or will they just be allowed to burn up on reentry without replacement reducing the swarm's capabilties?
IOW, is the basic idea too simplistic and reality far more nuanced, especially when technology is rapidly developing?
Levin: What is the model here? People invest in projects that have NPVs evaluated at externality-inclusive costs and benefits > 0 but with only private costs and benefit NPV < 0. A glorious outcome – exactly the opposite of what we normally worry about -- but why does it happen? Would it be better in these cases if the externalities were internalized? Should, could more such mistakes be encouraged? How?
Political correctness is the only possible explanation of Manhattanites not regularly discussing putting horses down.
Del Mastro: It is the case that Neoliberals have had to commit to lots of "affordable (with subsidies) housing" to get Progressives (who, just like "conservatives," see everything as zero sum) to go along with land use and building code reform. Just arguing that land owners, builders, city budgets, new residents and new businesses will benefit is not enough!
like many second best policies, the dose makes the poison. A low fixed percentage that does not have to be re-litigated investment by investment would be pretty harmless.
Levitz: Right conclusion but in need of two corrections.
a) We really cannot be sure if when and whose wages are rising faster than inflation. BLS produces unit value indices of wages, not price indices.
b) The concern about wages causing inflation is wrong not only because only the Fed causes inflation but also because it implicitly assumes that the only relative price stickiness that Fed engineered inflation needs to lubricate is goods/wages.
Rodrick: "Central bank independence" is a thing. It is possible that it be good (even a silver bullet) or bad. "Industrial Policy" is not a thing. There are a gazillion arguments for favoring one kind of firm over another, most bad, some good, but It just is not a thing that could be good or bad.
Of course Neoliberalism is a kind of ideology and I'm proud to show it.
You do realize that the failure of liberalism after WWI resulted in the 1920-30s shift towards fascism in Europe and Bolshevism in Russia. However proud you may be to be neoliberal, it won't help if history repeats itself and we end up as a fascist state in the USA. Idk what the solution is, but unless we can convince enough voters that they are not living in a socialist dystopia requiring a "strongman" to save them, then we could be in for a world of hurt (unless you are a white, Christian, heterosexual, male, and preferably with a family ancestry rooted several generations or more in the USA).