(Bloomberg Opinion) — There’s an inconvenient open secret behind many of the decisions we make every day. We live in an age of data. When we plan a trip, Google estimates our travel time and recommends routes based on current and past traffic data. Netflix suggests movies and shows we might like based on its data from people with similar tastes. The Federal Reserve uses market and macroeconomic data to make projections about inflation and what it should do about interest rates.
The open secret is that data makes us slaves to the past. And when the world changes, data doesn’t tell us as much about the future, and we may end up making worse decisions. Coming out of the pandemic has caused significant chaos and changed a lot of things in our lives. How this affects the data we use to make future decisions will be important for years to come.
To some extent, we have always relied on data to predict the future. Even when we were entirely analog, our past experiences – our personal “data” – determined how we perceived risk and made choices. But over the past century, computing power meant that everyone’s past experience, in the form of large datasets, began to influence our lives. This was especially true in finance, which took the prices of securities in the markets to make projections about future price movements and how to hedge or insure against risk.
Big data big problems
The data meant we could make decisions more scientifically, based on far more experience than our own. It offered great promise and more precision. Many aspects of our lives have become more predictable and efficient. Not only were we given the best route for our road trip, but our computer knew which shoes we wanted to buy before we did. Businesses could better forecast demand and spend less to maintain inventory, and airlines could sell more seats because they had a better idea of who would show up.
But this process has always been less scientific than we thought.
About 10 years ago I was tasked with projecting future interest rates for a retirement risk model. This problem posed many challenges. Normally in finance we take data from the past; when you look at a fund prospectus, it tells you its past average return and volatility. But in 2011, interest rates had been falling for decades, and there was a good chance that such a period would not happen again, because a repeat would mean that rates would turn very negative. Economists normally assume rates will return to their historical average, but there was no sign that was going to happen anytime soon either. The world had changed, it was more global and there was more demand for US bonds, which drove rates down.
All of these concerns led me to wonder if I could use historical rate data, and if so, what history? If I started in the 1980s, I would factor into my model a bull bond market that probably wouldn’t happen again. If I was using data going back to the 1970s, I would assume rates would go up – that probably won’t happen either. Ultimately, after much discussion with senior management, we decided to use as much history as possible and developed a process to update new data frequently. It was the most conservative choice, but it was a tough choice.
A living and breathing data stream
Data is no substitute for wisdom. Without good judgment, data will miss you because it binds you to the past. Assuming that house prices would not fall, simply because they had not done so for a long time, contributed greatly to the financial crisis. More recently, poor data judgment was one of the reasons the Fed was so certain that inflation was transitory. Fed economists use mean reversion models which, with the last 40 years of data, predicted that inflation would quickly disappear on its own. Changing data is one of the reasons why Google searches aren’t as useful as they used to be.
Today, the practice of letting data guide us has become a greater danger because our lives have become more data driven. Our online lives are generating more data than ever. And increased computing power and technology have enabled the use of this data. Suddenly, every trip we took, everything we bought, and every website we visited added to the predictive power of the data and guided our decisions – even if it was all a little scary.
But even powerful big data is a thing of the past. And now the world has changed and all that data isn’t so predictive anymore. Inflation is back and changing how and what we buy. Interest rates are rising, changing the way we save, invest and buy homes. And it’s not just the economy. The traffic patterns of 2019 – when people were commuting to work five days a week – aren’t as useful anymore, nor are the projections on the clothes we’ll need. We watch more movies at home and eat earlier. Leisure travel is on the rise, but there is still less business travel, making it harder for airlines to know how many seats to sell.
Today’s economy is chaotic for many reasons – primarily supply and labor shortages – so it’s hard to know how badly our data-driven algorithms are failing us. If the world has really changed, it may take years to have meaningful data to fuel our decisions again. Or, if it returns to normal, which pandemic years should we use in the future?
It’s unusual for the world to change so much in such a short time, but it offers a valuable lesson that we’re likely to forget: data is only a guide and never replaces our own judgment. Going forward, we’ll have to take all predictions, big and small, with some skepticism and balance them with our own personal experience.