A note on Wicksell
Issues to be addressed in this research paper
Over the years, the Swedes have delivered many things to the rest of the world that they can be proud of. ‘Made in Sweden’ has become a true stamp of quality.
Having said that, one of the very best ‘products’ to come out of Sweden is also one of the least well known. His name is Johan Gustaf Knut Wicksell, the most famous Swedish economist of all time. Wicksell was born in Stockholm in the early 1850s, and he was indeed very influential amongst his peers.`beta`
Now, almost a century after his death in 1926, Wicksell is still highly regarded, and his thoughts and ideas are applied by leading economists virtually every day of the year.
Wicksell operated with a very simple idea. At any given point in time, there are two interest rates:
1. The natural rate of interest (which we can only infer); and
2. The actual rate if interest (which we can observe).
He defined the natural rate of interest as the rate at which the economy is in balance; savings and investments are in balance, and the overall economy does well without overheating. Over the years, economists around the world have used the economy’s nominal growth rate as a proxy for the natural rate of interest.
The actual rate of interest is simpler – what is the cost of borrowing? The average company can borrow around the yield implied by the BAA corporate bond index, so the 10-year BAA yield is often used as a proxy for the actual rate of interest.
Wicksell’s theory has evolved into something economists and practitioners call the Wicksellian spread (Although most researchers have named it the Wicksellian spread, some just call it the Wicksell spread. They are two names for the same thing), which probably stands out as his most important contribution to the economic profession. A more in-depth assessment of the Wicksellian spread, and where we are in the credit cycle, is what this research paper is about.
At this early stage I should point out that Wicksell’s thinking is not only relevant to academics. As you will see shortly, the implications of Wicksell’s thinking are massive for those of us who are involved with investments every day of the year, and I would strongly suggest you finish this note, if you are not already familiar with his thinking.
The Wicksellian spread – what exactly is it?
In short, the Wicksellian spread is the 10-year BAA corporate bond yield less nominal GDP growth. Over the years, it has proven to be one of the best – probably the best – indicator of banks’ willingness to lend; hence it is a superb indicator of where we are in the credit cycle, and of what can be expected of economic growth going forward.
The lower the Wicksellian spread is, the more willing banks are to lend, and a negative spread has, over the years, turned out to be particularly reflationary.
In the early stages of a cyclical upswing, the Wicksellian spread is almost always near zero, and sometimes even negative. As the credit cycle matures, the spread usually widens, and the banks become more and more uncooperative.
As we enter a recession, the Wicksellian spread is quite high. In the 2001-02 recession, it peaked at almost 6%, and in the Great Recession of 2007-09 it exceeded 10% (exhibit 1).
Wicksell argued that capitalism works best when the cost of capital to businesses (i.e. the 10-year BAA corporate bond yield) is about 2% higher than nominal GDP growth, and time has shown that he was right. At that level, banks are adequately compensated for taking risk on individual companies.
The implications of a low Wicksellian spread
When the cost of capital is about right (i.e. about 2% higher than nominal GDP growth), corporates are incentivised to invest productively - i.e. to borrow for productive purposes. However, when the cost of capital is too low, as it was in 2004-05, 2010-12 and again in 2014-15 (see exhibit 1 again), corporates are incentivised to borrow to buy back their own shares. The Wicksellian spread therefore also affects the overall volume of buy-back programmes in the equity market.
The low US Wicksellian spread of recent years has had a detrimental impact on investments across the US corporate sector; the appetite for investing for productive purposes simply hasn’t been there. Consequently, when measured as a percentage of GDP, US corporate investments are now at the lowest level since World War II (exhibit 2).
Another problem with a low Wicksellian spread is that it drives up asset prices relative to income. Economic agents – both corporates and non-corporates – are incentivised to speculate, because capital is cheap to borrow, and we have just seen what effect that might have on corporate buy-back programmes, but the problem goes further than that.
Access to cheap capital also affects an asset class like property. People buy property which (a) is an unproductive use of capital, and (b) they couldn’t afford if the Wicksellian spread was anywhere near normal. That the rest of us are overrun by greedy real estate agents only make matters worse. Those of us who live in or near London will recognise this problem.
Yet another problem resulting from a low Wicksellian spread is that it keeps zombie businesses alive. Access to cheap capital allow zombie companies to survive another day – companies that under normal circumstances wouldn’t stand a chance, but they survive due to access to cheap capital. And because these zombie companies tie up capital that could otherwise be used productively, overall productivity levels deteriorate.
Finally, a low Wicksellian spread encourages business owners to run their business for revenue growth rather than profit growth (like Amazon). Why is that? When the Wicksellian spread is low (or negative), growth stocks outperform value stocks, so the business owner is incentivised to grow the business as fast as he can.
This is a problem from an economy-wide perspective, because corporates that convert to this business model have a habit of destroying the profitability of the industry they are in, just like Amazon has destroyed the profitability of the US department store industry. Hence a low Wicksellian spread leads to lower GDP growth.
By far the biggest problem caused by a low Wicksellian spread is massive amounts of misallocated capital. Way too much of it is invested either unproductively, or it is invested in companies or projects that are not fit for that capital in the first place, such as the zombie companies mentioned above.
The MacroStrategy Partnership LLP, which has done extensive work on the Wicksellian spread, has arrived at a measure of misallocated capital, which they calculate the following way: They subtract the Wicksellian spread from 2% and then multiply that result by GDP for the quarter in question. That result is again multiplied by-debt-to-GDP.
The result is there for everyone to see (exhibit 3). We are not (yet) at pre-2008 levels in terms of misallocated capital, but we are uncomfortably close, and we are way above the levels we experienced prior to the 2001-02 recession.
The Wicksellian spread in the UK
Allow me to switch gear for a few minutes. As we have seen over the last several pages, the Wicksellian spread is now rising in the US and is no longer reflationary, as it has passed the 2% ‘neutral’ line (see exhibit 1 again). As we have also just seen, substantial amounts of capital have been misallocated as the result of the very low Wicksellian spread over the past several years.
Now to the UK. Could one draw the same conclusions? Not quite. The Wicksellian spread here in the UK remains very reflationary and low enough to encourage banks to increase lending (exhibit 4).
It is therefore only to be expected that the economic cycle in the UK could run for a fair bit longer than the US cycle. We also know that the banking regulator in the UK is turning less and less hostile on UK banks increasing their loan books, as they have now largely completed the repair job after the financial crisis. So far so good.
Having said that, there is a ‘but’, and the ‘but’ relates to an increase in misallocated capital / non-productive investments (whatever term you prefer) – a phenomenon that is creeping in across the UK.
Take the UK property market – by definition an unproductive way to invest your capital. Due to sharply rising property prices in and around London in recent years, a growing number of people think investing in property is the best thing since sliced bread, and the buy-to-let market continues to excel despite various government attempts more recently to cool that market down (exhibit 5).
As far as the US economy is concerned, the higher Wicksellian spread is already starting to have an impact on loan growth. As you can see on the extreme right of exhibit 6 below, loan growth has definitely dipped recently and may soon turn negative.
As far as the UK economy is concerned, it will most likely go the same way as the US economy eventually (i.e. into recession) but, if you subscribe to Wicksell’s model (and I do), the UK is still several quarters behind the US in the credit cycle; hence also in the economic growth cycle.
One more comment before I finish. If you take another look at the amount of misallocated capital in the US (see exhibit 3 again), it is obvious that the Americans are already in recessionary territory. In other words, a US recession could be upon us any time.
Unfortunately, I have no data for the Wicksellian spread in the Eurozone, but would assume that the Eurozone economy is even further back in the credit cycle than the UK. A recession there any time soon is therefore highly unlikely.
29 March 2017
I have leaned heavily against Charles Gave from GK Research and Julien Garran from the MacroStrategy Partnership when putting this note together. Thank you to both of you for providing valuable input.