What kind of indicators may have predicted the upcoming financial crisis in the 2000 or 2008?
A Bit broad of a question, although here are a few ideas:
- Non Farm Payroll rate of change This is produced by the Bureau of labour statistics, and tracks total employed persons. Ideally one would seek the change in payrolls to view the health of the job market for a given month.
- large sample size.
- there is a bit of lag compared to the jobless claims indicator.
- Jobless Claims This is produced by the Department of labour and includes total applications for unemployment insurance.
- produced weekly
- very volatile and noisy
- GDP growth This is produced by the BEA and includes GDP growth with GDP defined as: GDP = Consumption + Investment + Government Expenditure + Trade Balance
- literally measures US economic growth
- produced quarterly there fore is a lagging indicator
- PMI (Purchasing Manager's Index) A measure of Purchasing manager's sentiment, generally speaking a PMI<50 is indicative of recession. The PMI is reported by the Institute for supply management.
- leading indicator
- very subjective reporting
I recently stumbled over an interesting study related to this question: Yaneer Bar-Yam et al. from the New England Complex Systems Institute (NECSI) published a study in 2011 that used "measures of collective panic" to "predict economic market crises". To cite a report about the paper:
[...] Research analysts have found [that] high levels of collective stock movements - or market mimicry - tend to precede crashes, which suggests that measuring the mimicry level of the market could provide significant advance warning of an impending stock market crash. [...] The researchers constructed a model of this mimicry to obtain co-movement data, which is the percentage of stocks that move in the same direction. [...] When substantially more than half of the stocks move in the same direction, this co-movement indicates higher levels of mimicry."
Personally, I think the results look interesting indeed, although it might be difficult to build an indicator out of these findings.
Anyway, the paper can be found at: https://arxiv.org/pdf/1102.2620.pdf