Interesting Google Trends - Economic Predictions

[Andy R]

I like to look at Google Trends every so often. Today I decided to plug in some economic buzz words and target them to USA searches

Here is one for foreclosure:
Google Trends: foreclosure

If you do the same search on recession:
Google Trends: recession

So if you do a search for "double dip" you can see what's happening:
Google Trends: double dip

Here you can see how many people are searching for "v shape" recovery:
Google Trends: v shape

It's pretty neat stuff to look at.

I downloaded the .csv for "recession" and made a 26 week moving average to see if you could identify a deviation that signaled things were about to go in the dumper. By the end of October 2007 it was clearly off and by January 2008 it was over 300% above the norm from the previous few years. IMO that's a pretty good signal things might be headed downhill and it would be proper to prepare.

I know. I know. Market timing is a game we should not play but it is interesting to think about how new sources of information could shed light on what people are thinking.

What other keywords would be good indicators of up or down swings in the markets?

 

[Meadbh]

There's a PhD thesis in there somewhere!

 

[Andy R]

I just read this and thought of my old thread here:
How Google Could Easily Become A Dominant Investment Manager - Forbes

 

[inquisitive]

I'm not sure that this data means anything. I looked at the trends for "recession" in October 2007 and since that time the same level has occurred many times, without additional corresponding recessions. The data is all sort of normalized by google, so I don't know that we can say that it had extra significance in October 2007. By January 2008 the market was already declining. So it would be natural for people to start searching for a recession? If the downloaded .csv is not raw data, does it really make sense to do a moving average on it?

 

[imoldernu]

It's pretty neat stuff to look at.

Thanks for posting... looks to be an interesting exercise, not just for finances but almost anything. Depending on the subject, coupling online interest with potential breakouts of anything from fashion to dog food, could obviate much dedicated lab and research work. My son works in a business that couples testing with analytics to provide guidance for advertising and product development, all done in a dedicated lab type setting. This looks to be an excellent source for valuable info and maybe some shortcuts.
.

 

[aim-high]

A friend of mine owns a polling company. They did some work on economic indexes and I know he spent some time looking for correlation between economic attitudes, employment, and the market.

I guess the big question is can you find correlation that is a leading indicator?

 

[clifp]

Thanks for posting... looks to be an interesting exercise, not just for finances but almost anything. Depending on the subject, coupling online interest with potential breakouts of anything from fashion to dog food, could obviate much dedicated lab and research work. My son works in a business that couples testing with analytics to provide guidance for advertising and product development, all done in a dedicated lab type setting. This looks to be an excellent source for valuable info and maybe some shortcuts.
.

There was book and Google talk that study sexual desire using Google searches. I guess I should point out that subject matter is NSFW Although no nudity is involved.