D2D Cont'd - Is the analog cycle turning?
Maybe, but it's hard to see in the data. So, maybe?
Is the Analog Cycle Turning?
The analog sector is notoriously difficult to track. Most analog companies sell into a dozen different end markets. Product cycles last years, sometimes decades. Design wins are hard to track, often only a few people at the customer actually know the analog content in a product. As a result, analog stocks tend to be tracked alongside broader macro-economic terms, and usually trade with their inventory cycle.
We launched coverage of Analog Devices and Texas Instruments with Sell ratings. We think both are good companies, but our thesis holds that the analog inventory cycle is not positioned to improve soon. It may have bottomed out, but we do not think it will turn positive for several quarters. Not everyone agrees with this. Chief among them, the analog companies themselves. During last quarter’s earnings season, several companies specifically said the cycle had turned.
Here we want to gauge this by looking back at the past five years’ inventory trends across a range of analog companies. Then we will walk through some of the major end-markets.
Looking at all this, it is hard to find a clear case for improvement. The inventory numbers do not look particularly inspiring. Nor do the end markets. As much as Aerospace/Military and some pockets of industrial look to be solid, other markets seem largely lackluster. In particular, we want to call attention to China’s electric vehicle (EV) sector. This has been a bright spark of growth for many analog companies over the past year. That may be changing. The PRC central government is now pressuring those companies to reduce their working capital, a process which we think will result in significant inventory drawdowns on their own books, which will leave stock in the suppliers’ warehouses, driving up their inventory. And of course, looming behind all of this are the US tariffs. While not as bad as originally feared, their impact has yet to fully percolate through the industrial base.
Cash Conversion Cycle
To better gauge the situation, we looked back at five years of working capital at nine analog (or adjacent) chip companies:
Analog Devices
Indie Semi
Infineon
Lattice
NXP
On Semi
Renesas
ST Micro
Texas Instruments
We calculated their cash conversion cycle to gauge their overall working capital trends, in the chart below. Eight of these actually saw their cash conversion cycle increase in the last quarter. Of those, five saw declines in the prior quarter. Glass half full – these companies are bouncing along the bottom before taking off. Glass half empty – false dawns are possible.
Cash Conversion Cycle
We then took a further step, looking at inventory days. Here the trend is very similar. Lattice and On Semi both saw days inventory decline, especially On, but the others all saw increases.
Days Inventory
In fairness, one quarter does not make a trend. That being said, these companies continue to hold historically high levels of inventories, and those show few signs of declining.
Segment Overview
Analog companies sell into almost every end market, there are analog chips, unnoticed everywhere around us. Among these, six drive a large portion of revenue. Below are quick thoughts on those six.
Automotive is the largest segment for most of these companies. Here the outlook is not particularly strong. To be clear, we are not experts autos, but to our non-expert eyes the outlook is not great. US auto inventories fell in recent months, but we believe some of that can be attributed to pull-forwards ahead of tariffs. There are also a lot of moving parts here, with US auto makers generally overhauling their component inventory practices since the pandemic. Our conversations with auto semis professionals point to a generally mixed outlook. End demand for cars is stable, but that is not translating into booming sales of semis. US auto makers have generally scaled back their EV plans, and moves towards new analog chip architectures (i.e. zonal systems) are moving at a glacial pace. There is nothing wrong here necessarily, but there is also nothing to make us think growth in semis demand is about to surge. The situation in Europe is a bit more muted. Auto inventories are up a bit, as carmakers there are struggling with a deluge of China EVs. Again, the outlook is not strong.
Probably most concerning is the outlook in China. This has been a major source of growth for analog semis for throughout this down-cycle. That may be about to change. The PRC government has recently pressured China’s auto OEMs to start paying their suppliers on time. Average days payables for EV makers is close to 100 days and in some cases over 200. PRC EV makers have somewhat peculiar. Unlike traditional car makers, they have very little debt on their balance sheet, but they do have massive working capital. They are effectively borrowing from their suppliers. The government is now pushing these companies to get payables down to 60 days. For reasons beyond our scope here, PRC companies often have significant cash flow constraints, with perennial problems of long debt chains (vendors cannot pay suppliers who in turn cannot pay their suppliers). Historically, this has lead to a lot of instability. Seeking to avoid a liquidity crunch, we believe the regulators are stepping in before conditions get too tight. What is not clear is how the EV makers are supposed to fund operations under these new policies. Absent a massive infusion of bank debt, we think the most likely outcome is for the suppliers to bring down their parts inventories sharply. The effect of this on analog semis will be a major headwind in coming quarters.
Aerospace and Military is probably the one end market with the best outlook right now. Aside from the demand created by the grim state of the world, the US Defense Department seems to be moving quickly to open up new ways to purchase advanced systems and weapons. All of these rely heavily on semiconductors, especially analog chips. The drawback is that quick by Pentagon standards is still pretty slow by semiconductor standards. This is market is also very fragmented with hundreds of small vendors selling into it. Good growth, but not a huge market on its own.
Telecom may have bottomed but it is not going to grow any time soon. Outside of new data centers, there are few growth catalysts in networking. Data center needs are small relative to the rest of the segment, so are unlikely to alter the overall trajectory. The 5G build is over, and 6G is still years away.
Healthcare is another area where we are not experts, but we think the US government push to cut spending – both on patient care and for research – are going to wreak havoc on demand from this sector. On the pharma side, we now that funding levels are very tight, with a growing number of layoffs for research and manufacturing staff. We know that demand for manufacturing equipment in this sector, which requires significant semis content, is slowing dramatically amidst the broader changes in the industry.
Building Automation is unclear. Housing construction in many markets seems to be generally stable, but construction of commercial and industrial is not. Here we think the tariff impact still has the potential to be a major factor with many inputs now considerably more expensive.
Test and Measurement is a mixed bag in its own right. T&M systems for semis are very strong right now, but demand in other markets mirrors the varied conditions above – healthcare dim, industrial modest.
We could go on, but the point is clear. Other than aero/mil, n segment gives much conviction of strong demand for semis. For every market that looks decent, there is another one that looks gloomy.
Behind all of this are two big, related questions. What will be the impact of US tariffs? And will the US enter a recession? Neither of this look as scary as they did a month ago, but neither do they inspire much confidence.
Conclusion
Between the lagging indicator of inventory levels and forecasts for end markets, it is hard to build a clear case for sustainable improvements in industry conditions. Our view is that the sector has at best bottomed out. And while we see no reason to think there is another leg down, we do not see anything that would give us confidence growth is imminent. Given the high degree of uncertainty around the economy, we think a more likely scenario is that a recovery in the analog cycle may still be several quarters out.
Like always, invaluable insights