11 ms·
I'm not gonna put out my own model, but this one seems strangely, suspiciously well fitted to the historical data. How can you have a model that gives a 100% c
by lordnacho 7y ago
I'm not gonna put out my own model, but this one seems strangely, suspiciously well fitted to the historical data.
How can you have a model that gives a 100% chance of something happening in the future? Yet that's what it says during the last recession. And it falls off a cliff when the actual recession ended in June 2009.
I was alive and trading during that time, and I reckon the business cycle would be quite easy to predict -as in, the model would be in a textbook- if you could be that confident of recession one day and the opposite the next.
- hogFeast 7y agoI think the 100% is when the data shows the recession is occurring i.e. if a recession is occurring now, then the chances of one occurring in the next 3 months are clearly 1 (I get your point however...it is odd). There are other factors in the model aside from the ones with charts. The model isn't in a textbook because predicting the economy is nothing to do with economics as it is taught at universities. Economists use complex structural models because they are fun to teach and create barriers to entry. This is why DSGE models are so popular in universities, govt, and central banks but barely used anywhere else. If you want to forecast the business cycle, you just need linear regression. You can't ever be 100% (unless a recession is already occurring) because the economy is always changing but basic models are pretty accurate (the BIS has done quite a bit on this recently - https://www.bis.org/publ/work818.htm https://www.bis.org/publ/work818.htm - funnily enough, the term spread isn't the be all...but economists have a big problem with including things like credit.