Ode to localization?

Sometime last week when Indus Motors (PSX: INDU) submitted its financial statements to the PSX, BR Research predicted that the company would raise prices soon. We did the math. We made a simple estimation of the revenue per “assembled” unit sold and found it rose 55 percent during FY23 (versus volumetric decline of 58 percent), while the same increase in costs per assembled unit sold was higher—at 58 percent. This caused margins to decline. To keep margins steady and looking up, prices will have to be raised.

One could question how costs rose so much more than the rupee depreciated during the period—which was about 40 percent on monthly average basis during the fiscal year. It would be a lot less if the first working day of FY23 is compared to the last working day of the fiscal year but since inventories are replenished at various times during the year, it makes sense to take the average for the year, rather than the spot rate. Nevertheless, price hikes certainly exceed the rupee depreciation by all accounts—51-58 percent for various models between June of this year and last. Because Indus Motors increased prices in July-22 and then reduced them by 7 percent in Aug-22 and then raised them again in Nov-22, perhaps, the better estimate here would come from taking the average of prices during FY23 and FY22. This shows a price hike of 63 percent- on average.

Earlier, BR Research estimated that the average CKD per unit sold in the country during FY23 (in rupee value) stood at Rs1.6 million (Rs 16 lakhs) per month, which was up from Rs1.1 million last year (read that story here: “Autos: The scatter here is too great”).In other words, during FY23, the industry imported Rs1.6 million worth of CKDs per unit of sale. Standard warning that is a close approximation, rather than exact numbers. For instance, only volumetric sales for PAMA members were used here. If non-PAMA members such as Master Motors or Kia were included, the per unit CKD cost would belower.

The company claims that a major portion of the price goes to the government in the form of taxes. Let’s assume this to be accurate (for the purpose of this story). This would somewhat explain the huge difference between current prices and average CKD costs—today’s cheapest model of Corolla costs a whopping Rs6.1 million. Sans taxes (assuming a tax of 40% on this model of Corolla), the car would roughly be priced at Rs3.66 million to consumers. If we were to take an average price for all Corolla models, it would come around to Rs5.9 million. Let’s extract 40 percent tax from this price, the pre-tax price would stand at Rs3.54 million. That’s a big reduction indeed. If Indus Motors takes a gross margin of 5 percent on Corolla, the localisation level comes to around 50 percent (using the CKD cost estimated at Rs1.6 million). If the gross margin is 10 percent, localization should be around 45 percent. If one were to further include estimated sales of non-PAMA members (who are also importing), the CKD per unit would be less—somewhere around Rs1.3 million which would show a greater localization level of 58 percent. We continue to assume here that the company maintains a gross margin of only 5 percent on this vehicle (which is likely not the case). There is a lot of guesswork here since it is nearly impossible to estimate the CKD cost per unit incurred by Indus Motors based on the information available publicly. Essentially, we could place a ballpark around 45-55 percent localization level.

But we have yet to stir the pot. For that, head over to Indus Motor’s Annual Report for this year. On its second page, against the backdrop of a ‘radiator assembly with fan’, three big bold letters claim “Make in Pakistan” and at the bottom, the report proudly declares, “an ode to localization”. Indus Motors claims its local-to-import ratio is 60:40 for Corolla and Yaris (the ratio is flipped for Fortuner and Hilux). Not to be a stickler when we are using guesstimates, but this is higher than our calculations. But then the company says there is a 90 percent chance if you randomly touch a Toyota vehicle, you are touching a part made locally, and we have a problem.

This is only true because Pakistan’s existing localization capabilities do not extend beyond making body and moving parts. How localized can the Toyota car in Pakistan be if every sneeze the rupee takes, the company catches a cold. After all, localization is expected to reduce the import bill, right? Actually, what we keep referring to as localization is not “local” content or value-addition, because most Tier 1 vendors of OEMs are themselves importing materials and components in dollars to assemble parts. Our generous calculation of 58 percent localization level is not local content.

It’s true that price-setting behavior in Pakistan – broadly –is influenced by the movement of PKR. It may also be true that assembling vehicles at home implies jobs across the value chain. But is this the most efficient allocation of our resources? Could we have employed these people in sectors that could become more competitive faster? What is the level of technology and knowledge transfer in Pakistan that has happened over the years with auto parts vendors benefitting from foreign partnerships? Pakistani vendors typically work with OEMs and their existing blueprints and designs that the OEMS get from their parent companies. This to a degree restricts vendors in their learning progression. There is a reason why Pakistani auto parts vendors are not making or investing in high-level technical parts. No one would buy these parts from them.

We have an industry here with criminally low volumes compared to the population it is meant to serve—in FY22, government data reveals there are 4.6 million motor cars on the road (including jeeps). Assuming an average household size of 6 members, car penetration is 12 percent of households. But at the higher end of the income scale and in rural settings, there are more cars per household, so penetration should be even lower.

This is where we stand. Latent demand for automobiles is enormous but there hasn’t been enough supply. Market size is stagnant. Capacities have remained restricted which have historically caused prolonged delays in the supply of vehicles. A PIDE research found that 90 percent of the vehicles sold in Pakistan are on “own-money” or sold at premium. Why? Because supply hasn’t kept up, but demand is there. The benefits that could be accrued from economies of scale are significantly marred by the lack of commitment of OEMs to invest into new capacities, and frankly, models that would capture the desired volumes.

At the same time, auto-development policies purporting to induce localization have been a colossal failure. There has been no focus on developing the raw material industry (steel plates, plastic resins, etc.) or truly deepening the technical capabilities of domestic vendors and manufacturers. Critically, the work on safety and quality standards that would bring uniformity in product development at par with international standards has lagged behind. Ideally, the decades-long protection the industry has enjoyed should have expanded volumes for each model introduced, and made necessary investments across the automobile value chain, eventually leading to exports that brought in dollars. None of this happened. Industry players remain insulated from any real competition—be it CBU imports or used car imports, only making cars for a certain class of people. By no means, this is an indictment of Indus Motors as a company alone. Shoddy governance and poor, short-sighted, unimaginative policymaking certainly stand front and center on the blame scale. But in FY23, Indus Motors raised Corolla prices 6 times—by Rs2.7 million on average—and then patted its own back on “making” its cars in Pakistan. Here’s the thing: when rupee is sneezing, it’s not acold being caught, it’s pneumonia. And it’s certainly not an ode to localization.