Saturday, September 17, 2011

Log of (jim's graph)


Jim's graph, from my four o'clock:

Graph #1: Change-From-Year-Ago in Financial Debt, billions

Log of (jim's graph):

Graph #2: Log of Graph #1
The three peaks are still visible on Graph #2, but the overall trend is rather a straight line. Couple big whoopsies at the start. What else? The big decline at the end of Graph #1 is minimized here.

Some truth to the rumor that growth is exponential. When you find exponential trends everywhere, it loses some of the excitement. Still, I like what Clonal said about total debt yesterday:

If you deflate [total debt] by population growth, the cpi, and a real gdp growth index, you will still find that there is residual exponential growth...

In other words... Population aside, prices aside, and economic growth aside, our use of debt has still increased at a remarkably regular rate.

Yes. Which reminds me of something Liminal said:

The New Arthurian Economics blog, which involves a retired guy constructing lots of wonderful charts from FRED data and then using them to argue that we have too much debt and too little money...

Yes. Which reminds me of something else... Suppose you take two datasets that definitely have exponential growth in them. And you divide the one by the other. So you are sort of dividing the "growth" out of the numbers.

What does it mean if the result is exponential when you graph it? It means there is some constant force other than growth pushing in one direction. Clonal finds it, after stripping inflation and population and economic expansion away from total debt. I find it in my Debt-per-Dollar graphs. "Too much debt and too little money".

There is some constant force there, acting to increase debt. I think that force is human nature, expressed in economic policy.

No comments: