![](https://www.cuemacro.com/wp-content/uploads/2022/10/20221001-Importance-of-forecasts.jpg)
Sometimes we can really predict the future. These days weather forecasters ingest massive amounts of data to come up with their forecasts. Admittedly, they are not always 100% accurate, but they are noticeably more accurate than they were say 20 years ago.
In financial markets, the likelihood of being extremely accurate in predicting the future is much lower. However, thankfully the threshold to being profitable is much lower. If we can predict the direction accurately just over 50% of the time that is hopefully sufficient to being profitable (ok, I know it depends on the skew of the returns in practice…). Although forecasts cannot be as accurate as in other domains, they are nevertheless a key part of the trading process.
Market participants create their own forecasts to guide their investing process. Independent forecasts compliment your own forecasts. The machine learning based inflation forecasts we produce at Turnleaf Analytics, the firm Alexander Denev and I co-founded, are often used by clients in tandem with their own views. If we have no forecasts at all, it is as though we are driving without any headlights. Headlights cannot illuminate all the path in front of us, but they are do help.
This week gave us an illustration of why independent forecasts are so important. Part of the market turbulence in the UK was related to the lack of independent forecasts from the OBR after the Chancellor’s mini budget. It isn’t that OBR forecasts are always 100% accurate (!), it is that they provide an independent viewpoint of his spending plans. Independent forecasts can sometimes tell you something which disagree with your view, but this makes them even more valuable, providing you with another key data point and a way to judge your own views. Independent forecasts are obviously most valuable if they are more accurate and better than the market benchmark.
Indeed, for Turnleaf Analytics inflation forecasts, we have created additional models which are representative benchmarks of the market, as a way to understand how our forecasts compare to the market. So far, in our live publications, our forecasts have beaten the benchmark in 59% of occasions.
Forecasting is important in financial markets, even though we know it can never be as accurate as other domains. Even if we have our own forecasts, having a good independent accurate forecast can be an important data point to incorporate for our own decision making. Having a benchmark can be a good way of evaluating the quality of a forecast too.