indianotes

Humans don’t like uncertainty. And nowhere is this more evident than in economic or business forecasts. Our analysts and economists cater to this need, with a great degree of apparent precision and accuracy.

These forecasts further reinforce our belief that such things can actually be predicted.

Consider these statements:

• “We estimate consolidated PAT to grow at 39% CAGR over FY25-FY28E.”
• “We cut our FY28E EPS by 1%, mainly owing to…”
• “The two-wheeler industry is expected to grow at 7% over the next three years.”
• “The capacity increase will result in a 2% increase in market share for the company.”

Sure. As long as no pandemics, wars, or alien invasions show up.

Is this for real? The past decade has featured demonetization, GST implementation, Covid, large and small wars, supply chain disruptions, AI revolutions – and now Trump – and we still think we can predict 2028 earnings down to two decimal places!

Even crazier, everyone just accepts such accuracy as par for the course. In fact, it is often the user/customer demand for precision that drives analyst behavior, and not the other way around.

Many years back we were trying to model future auto sales for an ancillary manufacturer. Long term data clearly showed an upward long-term bias, coupled with substantial cyclicality (booms and busts). But they insisted on getting a single number, year by year. We refused and instead developed a set of alternate scenarios that would help predict the cycle, rather than a fixed set of numbers. Our approach was correct, but we lost the client.

Another time our client needed to show a higher industry growth rate to justify his budget requests. His boss wanted a lower number. We had provided a range, but they both insisted on a single number – each closer to their own vested interest. This too, didn’t have a happy ending.

If an equity analyst were to say, “I cannot predict next year’s EPS, but I think you should buy …,” he or she would probably get trolled by investors.

On almost every analyst call I’ve attended, company execs get badgered for “guidance”. Since they can’t say, “I don’t have a clue”, the normal reply is, “We don’t provide guidance.” Yet, analysts continue to ask for exact data or assumptions (that they can feed into their spreadsheets).

Many institutional brokers and funds build amazingly complex models to analyze and forecast financial performance. We once worked for a fund in helping populate and update these models. There were several connected excel sheets, innumerable assumptions to be made and hours of research to find all the data; that simply populating a single model took three weeks!

Every number, ratio that you can think of, and several that you can’t will be calculated. This is despite the fact that a single wrong assumption (or Trump post) means the whole thing is anyway, useless. But if they don’t do this, their clients will somehow think less of them.

Maybe this is simply a ploy to keep analysts employed, or maybe a giant conspiracy by Microsoft Excel?

The reality is that the world, economies, industries and companies face an enormous amount of uncertainty. While a certain trajectory can be predicted, or broad comparisons can be made, it’s simply impossible to be precise about future growth, earnings or share prices.

Yet, business leaders, fund managers and investors can and do take the right calls. They do this by accepting uncertainty, and trying to manage risk – rather than depending on a precise view of the future.

In business and investing, as in life, knowing we don’t know might be the best forecast of all.