Discover more from Gad’s Newsletter
Is Nearshoring Finally … Near?
In a recent survey, 1,610 executives in Europe and the US were asked about their capital spending plans in the aftermath of COVID, the recent supply chain issues, and the war in Ukraine. A third of the respondents said they will look into nearshoring new operations, while 37% said they plan to return production to the US.
Is it finally happening?
A few weeks ago, I wrote about how at the beginning of the COVID pandemic firms overestimated the extent to which they were planning to nearshore their production.
This graph is based on a survey conducted by McKinsey over a year ago (and admittedly, with a much smaller sample size), but the picture remains quite clear: rather than nearshoring (either the production or the supply base), which 55% of firms were planning to do, most used other tools to deal with the situation, with the majority just increasing inventory. And as we now know, they are all currently sitting on a lot of that inventory.
But the intention to nearshore is not a new trend.
We have been hearing that the next phase will involve nearshoring since 2010:
“After a decade of rapid globalization, economists say companies are seeing disadvantages of offshore production, including shipping costs, complicated logistics, and quality issues. Political unrest and theft of intellectual property pose additional risks.”
“Caterpillar said Thursday it could triple its domestic output of construction excavators by consolidating production from an existing factory in Akashi, Japan, and one near Chicago to a new U.S. plant whose location is yet to be determined.”
So let's try to first understand the trend.
Some seem surprised to learn that it started long before the Trump era and COVID, and even before supply chain issues began.
And the main reason was the fact that manufacturing costs in the US moved closer to levels found in the main manufacturing hubs, as the latter became more expensive. Primarily, as the middle class in China, for example, increased in size, they expected higher wages. In 2015, the New York Times published an article stating:
“Manufacturing costs rose from 2004 to 2014 in most of the top 25 exporting countries. Some that are usually thought of as cheap places to produce goods like Russia, Taiwan, and China now have costs pretty close to those in the United States.”
And the next graph illustrates this well:
Admittedly, the trend has not only been about bringing manufacturing back to the US, but also (and probably mainly) to Mexico.
As ‘proximity to markets’ became important, Mexico became more viable. Even when the cost of production was not the main concern, the difference in the Total Landed Cost (the end-to-end cost, including transportation, duties, inventory, etc.) was significant.
In general, the two main factors that drive supply chain and operations are scale and uncertainty.
As scale (and efficiency) became increasingly important, the likelihood of products remaining offshore increased too. But as demand volatility, and responding to consumer demand began to matter more, nearshoring became more interesting.
Note also that volatility is not only about demand itself, but also about what we refer to as “effective volatility.” In particular, the more products you offer, the higher the effective volatility (since it’s harder to predict each product’s demand). On the other hand, the shorter the lead time you promise a customer, the higher the implied volatility (since it’s harder to forecast when customers expect an immediate response), meaning you are more likely to move nearshore.
And indeed, as Amazon seems to be forming the consumer trends and dictating the rules over the last few years, we inevitably see other firms following suit and both shortening lead times and increasing the number of SKUs they offer.
During the Trump Administration, the nearshoring trend increased further, specifically with respect to China when a new set of tariffs came into effect. As you can see in the graph, taken from an AT Kearney analysis, China’s loss was translated into gains by other countries (some for Mexico, and some for the US):
But looking at the bigger picture, we see that the US Manufacturing Import Ratio (MIR) has only been increasing, which means that the US increases the amount it imports faster than it increases the amount it produces domestically:
Note that the constant increase in the MIR doesn’t exclude the fact that maybe more firms moved to be in Mexico, but it does show that nearshoring is not actually the massive “trend” they are setting it out to be. Another way to put it: It’s possible that we produce more in the US, but we also consume more, and the consumption increase is so much higher than that of production, that we continue to import more, and also increase how much we import.
The last two years have been so tumultuous, that I’m not sure we can learn a lot from them regarding trends, but until 2019 (pre COVID), we were seeing a slowdown of the nearshoring trend and, at one point, even a reversal where the ratio actually decreased (and not only increased slowly).
So it still begs the question, why is the trend so much slower than what we expect? The belief in 2010 was that this was the trend, but from 2015 until now, we have only seen very slow progress.
So what can explain this?
One possible (and evident) reason can be that it’s hard to find a new country to produce in (or even find suppliers in). It may seem like a turn-key operation, but the reality is that moving to a new country requires significant capabilities, data collection, and a level of project management that most firms do not have. Managers that don’t have relevant experience with such projects tend to overestimate the simplicity and speed of establishing operations in a new location. Specifically, during COVID, it was hard to travel in order to find good suppliers. Even with the most advanced software, it's hard to do business with physical products unless you meet people and see the plant, the infrastructure, and the managers.
But I think the reason is actually deeper. To paraphrase Erykah Badu, ‘we are linear people in a non-linear world.’
Things are going to get mathematical.
Several years ago, I co-authored a research paper with Jan Van Miehgjm on dual sourcing, balancing near and offshore production. In this study, we compute the amount that a firm has to source from its offshore location using the following equation (for those who are curious):
Where h, is the inventory holding cost per unit per period, Δc is the offshore sourcing cost advantage per unit, and σ is the standard deviation of demand per period.
So the fraction of demand a firm should order from its offshore location decreases linearly with the volatility of demand (or supply), but decreases only as a square root in terms of the cost difference between the near and off-shore location. Volatility has a first-order effect, while cost only has a second-order effect.
One important implication is that firms tend to think that they need to nearshore faster than they actually do. A 20% increase in cost in the offshore location, doesn’t translate into a 20% reduction in the viability of offshoring. So my hypothesis here is that firms extrapolate from the changes they see in costs to what they need to do, but when they actually do the math (assuming that some math is being done), they realize that they don’t need to move so fast.
Of course, this is only a hypothesis.
So what’s next?
While the long-term trend will be toward more nearshoring, just like everything in OPs, it'll be slow, and for a good reason:
Interest rates are increasing. The world is becoming more volatile, with more weather-related issues, more geopolitical issues (Ukraine is only the first step. Is anyone here willing to bet that China won’t invade Taiwan in the next 5 years? I’m not). More cyberattacks, more pandemics, and more volatile demand (with the economy already on a roller coaster). And again: the only thing with first-order effect is volatility.
Thanks for reading Gad’s Newsletter! Subscribe for free to receive new posts and support my work.