Semiconductor stocks have had a bad month. Nvidia was down nearly 10%+ in September and has fallen nearly $50 from its all time high.

It isn't alone. US semiconductor companies have taken a beating or a breather, last month depending on how you look at it. See here:

This raises the obvious question of: fine, they were probably due for a tough stretch but what now?

Well, we have detailed the macro issues above and we would expect semiconductor chips stocks to move in tandem with oil price and interest rates for a bit.

But what about specific drivers of the sector's profitability? Are there not any themes that will determine, positively or negatively, how they do?

Well, one of the optimistic takes on chip stocks is that there is money flowing into the sector - including huge amounts of cash from governments all over the world. These government subsidies, the argument goes, will pad chips' company profits for many years and be a new and huge source of revenue for the sector.

We aren't sure.

Now that is hardly surprising perhaps. This newsletter has been highly critical of the new political passion for old school industrial policy here in the US.​

In short, we think it most of these policies will be ineffective, wasteful and miss addressing the actual problems they are trying to solve.

The best summary for our view might be that the infamous "Inflation Reduction Act" will likely cause higher inflation with its many, many domestic buy provisions, regulations and complexity.

Our take isn't a popular view these days to put it mildly but we had also qualified it slightly. The exception to our argument was, we hoped, the US Chips And Science Act. We had held out hopes that actually, while unpleasant, this more focused piece of legislation might be worth the substantial investment by US taxpayers.

As a reminder, the reason we were more supportive of this was two reasons:

1. There was a clear strategic imperative for the US to have some of the most sophisticated chip design and especially manufacturing done domestically.

2. We hoped that at least some of the capital would be devoted to research & development not just incentivizing factories to return to the North American continent.

On the former, it might be expensive to bring some highly sophisticated manufacturing back to the US but strategically it would be worth it to spend some extra dollars to have both the supply chain and the know-how located domestically.

On the latter, industrial policy has a long and proven track record of failing across the board with one very large exception: R&D.

Semiconductors are the ultimate item in the 21st century arms race and so devoting more capital towards developing the next advances in this critical field would be very worth it.

Sadly it is time to admit we were likely wrong in this view. The US Chips Act is now coming closer and closer to being a similarly flavored waste of capital as the incredibly misnamed "Inflation Reduction Act."

Why?

Two reasons:

1. In February the Commerce Department announced that any financial support would come with certain conditions attached.

2. Now it appears that actually we are also dramatically expanding the scope of what industries can receive access to the subsidies.

Both of these are very bad signs for the utility of the Chips And Science Act.

On the first point we are foolishly prioritizing many other often conflicting goals besides simply bringing back the most sophisticated manufacturing back to the US.

That means it won't happen. Or it won't happen much.

This happens a lot with worthy liberal initiatives. The impulse comes from a good place and the individual goals are worthy in an absolute sense but because you end up trying to accomplish too many goals - regardless of whether any are good or not - you accomplish none.

It is a classic "jack of all trades, master of none" type scenario.

it isn't just cynical newsletter scribblers that have noticed this problem either. Even some of the high priests of liberalism have realized that in righteously trying to cure poverty and injustice and climate change and drug addiction and mental health crises etc etc all at once you often end up achieving nothing much at all.

Here is Ezra Klein in the New York Times writing on the subject. He amusingly calls it "Everything Bagel Liberalism." The idea being that by trying to add in all these different policies you just achieve policy goop rather than a tightly focused and likely more achievable goal. That isn't much like a delicious everything bagel in our opinion so the analogy might not be great but the phrase is at least memorable.

When it comes to the Chips Act, by trying to get the semiconductor companies to sign up for endless other progressive goals - worthy though they may be by themselves - you risk accomplishing much less and perhaps not what you intended in the first place.

What goals:

• "Buy American" provisions for one.

• Mandating using union labor for two.

• The Biden administration's "Good Jobs Principles" for third.

It is worth pointing out also that none of the above were actually in the Chips Act legislation itself. So, no one voted for it but that is second to the fact that most of the goals for these funds are not actually about advanced semiconductor chips.

i.e.: you wanted very advanced, cutting edge semiconductor plants to act as as a strategic bulwark against Taiwan's geopolitical vulnerability. That was the hope. But instead you will have middling plants with lovely social facilities that produce useful but mid market silicon semi chips at exorbitantly high prices that are made possible only thanks to the lavish US government subsidies.

It doesn't that THAT cynical a mind to sadly realize that the former scenario is not just rising in probability but likely the base case.

On the second point, just recently it became apparent that actually this wasn't about semiconductor companies alone.

That is because the US government has decided to expand its subsidy regime to suppliers of semiconductor suppliers as well. So, chemical firms, tool manufacturers and producers of raw materials for semiconductor chips will also be able to apply for the funds made available by the Chips Act.

This is bad. It is one thing if you spend the scarce monies poorly on non-strategic goals and outcomes but hopefully at least some of it will end up being spend on the strategic imperatives of having semiconductors built domestically. Now however, some of these precious funds won't even be going to strategically important firms at all.

Quite why this expansion of the subsidy regime is happening is a but unclear.

Is it because suppliers won't relocate operations to the US or is it because the Biden administration have realized that they can use a back door to influence a far wider swath of the American economy?

It is likely a mix of both but it seems very probable that the bureaucrats at the Commerce Department have discovered just how few plants and facilities will really be involved otherwise and decided to expand the remit of their brief.

There is a lot we can say here but let us just summarize by saying: it isn't great for the long term number of cutting edge semiconductor chips that will be made here in the US.

We are still clinging to the hope that some of these factories will actually be of strategic value and won't just be dedicated to building lower value basic semiconductor chips but even then, we aren't so sure.

Lastly, it is important to note that with all these rules and competing goals, these subsidies are not going to help these semiconductor firms make greater profits. That will mean investors will stay away. In fact, they may do the very opposite and flee the new popular sector. Subsidy regimes often turn vibrant and highly competitive firms into firms that primarily excel at fighting for scarce subsidy dollars.

Look no farther than Intel. Once an American champion of cutting edge semiconductor technology in its own right it has now become essentially a subsidy machine in the US and around the world.

They good at it! Look no further than the $10 billion dollar subsidy package that Intel's executives somehow negotiated out of Germany.​

This has been good for Intel's bottom line - and its share price - but it certainly isn't likely to produce anything advanced or ground breaking and therefore strategic. Like a long list of once great firms before it you can be good at innovating or you can be good at lobbying for subsidies but rarely both.

As we have said a few times here at Pebble, bigger is still better. Remember that going forward.

*******

Have questions? Care to find out more? Feel free to Download our App (!!) or reach out at contact@pebble.finance or join our Slack community to meet more like-minded individuals and see what we are talking about today. All are welcome.