Online Jobs Ads Indicate Weak Labor Market

I like to look at every type of economic metric in doing research. The current environment is somewhat unprecedented in a few ways. One of those ways is the jobless claims stats. The jobless claims 4-week average hit a 44-year low which makes no sense because the labor market has expended. Considering unit labor costs are increasing faster than pay growth because of declining productivity still doesn’t make up for the supposed tightness in the labor market. If the labor market was the tightest in 44-years, the weekly earnings growth should be the highest in 44-years. Due to the increase in the labor market size, there should be the highest weekly earnings growth in many decades. There is something wrong with the jobless claims data without a doubt. I try to understand the potential inaccuracies of it, but we won’t know until the next business cycle because we’ll have something to compare it to.

As I said, I like to look at all data points. It’s interesting to see a data point which disagrees with the jobless claims and the unemployment report because they don’t reflect the weak growth in wages. As you can see from the chart below, the number of total online job ads has decreased from early 2016. Judging by this chart, the labor market is weakening. I’m not saying I agree with the findings in this chart, but it is worth nothing. The Conference Board is reviewing the methodology of this report to see if it can be improved since it differs from the establishment report. It seems that whenever a report shows undue weakness, the reports are changed, but when they are too strong nothing is done. The Q1 GDP reports were showing weakness, so the methodology was changed. There isn’t much pressure to look into why the jobless claims are so absurdly low compared to the size of the labor market.

The chart below explains the point I mentioned earlier about how worker productivity is declining. It does an amazing job at showing how firms have allocated their capital in the past 37 years. As you can see, the Capex was 80% of operating cash flow in 1990. That is partially what led to the internet boom. Capex spending has declined since then. It’s now close to 40% of operating cash flow which means the productivity growth in the next ten years will likely be as low as it has been recently unless changes are made. Buybacks were barely registering in this chart in the early 1990s. The explosion of buybacks was experimented with in the early 2000s. This preceded the financial crisis. Buybacks didn’t cause the crisis, but they were terribly mistimed; they boosted stock prices, making the decline sharper.

As someone who started following the market in 2007, I was raised on the notion that buybacks were terrible because of the results in 2008 and 2009. It gave us the first taste of the negative consequences of them. In theory buybacks are a great way to avoid taxation and return capital to shareholders. That theory is being tried again which is surprising because usually mistakes aren’t repeated right away. The problem is agency risk. Management benefits from buybacks because they boost short term equity performance. Low interest rates help the trend along. Many investors actually cheer on an action which will hurt them in the long-run. Shareholders also like the benefit of short term gains.

The current picture looks identical to the picture in 2008. Debt issuance is high in both periods. Capex is low in both periods. Buybacks are slightly lower in 2017, but dividends are slightly higher than 2008. The spending on acquisitions is about the same and the total spending as a percentage of operating cash flow is also about the same. That’s not a nice comparison to give the corporate spending, but it’s unsurprising. Many indicators have been flashing a warning sign since 2016, but the same weak growth phase has continued.

The chart below shows the current length of the business cycle compared to the past few cycles. The recovery has lasted 95 months. That’s the third longest in the past 60+ years. I can see the recovery passing the one in 1960 for second place. Obviously, the Fed, the CBO, and the IMF will never predict a recession because they are establishment/government organizations who want to keep everything going well. In 2013 before the crash in oil, an oil company wouldn’t explain its potential for losses if oil fell. It’s the way incentives work. It gets wonky when the Fed or CBO put out projections for everlasting growth after the cycle has been growing for a few years because they end up projecting something which has never happened. As you can see, the CBO now is expecting the recovery to last 207 months.

This chart looks identical to the chart from a few weeks ago which showed how the VIX was below 11 for a record length of time. If the record recovery length was broken, it wouldn’t be such a great achievement because total growth is sagging. Having recessions followed by real economic growth spurred by productivity growth is probably better than a zombie economy with low growth and no recessions. The comparison between the economy of the past and the current one isn’t complete until we find out what type of recession comes with this new low growth economy. Will it crash like I think it will or will it have an equally shallow recession?


The fact that online jobs aren’t plentiful is interesting because working online is more popular than ever. The chart should look like online shopping where it goes up unless there’s a major crisis. I am in no position to judge how such an index is calculated because it’s likely very complex. I will wait to see what changes are made to it in the future. There would need to be a major change to make the index not show a decline. Either it was always an inadequate index or something changed recently which is throwing it off. The jobs market on the internet is ever changing, so that wouldn’t be a shock.

Spread the love

Comments are closed.