Introduction 

Global growth is fading, central banks are talking about cutting rates, and the US yield curve has inverted.

Now is a good time to think about how to invest during a recession, what a recession would mean for your portfolio and what strategy can you adopt to ride out the volatility?

What is a recession?

Before we look at how to invest during a recession, let's start by exploring exactly what we mean by recession.

The most frequently used definition is that it's two consecutive quarters of decline in real gross domestic product (GDP). GDP is the total amount of goods and services produced by a country.

In the U.S. the National Bureau of Economic Research has a slightly different definition. They say that they don't just use real GDP they also use a range of other indicators, particularly indicators which arrive every month that allows them to create a more timely definition of when activity starts to decline. They also consider how large the decline is and the phrase they use for this is "a significant decline in activity".

The Bank of England has data on UK growth going back all the way to 1700.

A big cause of recessions in the past was war. In the graph below they have shaded the major war periods in blue.

From this graph we can see that before the early 1800s we spent about half our time in one war or another and as a result we had this continual onslaught of recessions leading up to the early 1800s.

Recessions before the 1850s also depended on things like poor harvests, because the UK economy used to depend more on agriculture. There were also investment fads, for example investing in canals or toll roads. There were also very large and significant bank failures and you can see that after 1850 the frequency of recessions dropped quite dramatically.

The First World War and the Second World War also had their recessions. Most recently the most significant recession was the 2008/2009 financial crisis.

How long is a recession?

Given that we have this very detailed history of recessions in the UK, we can use that data to see how long a recession typically lasts. 

In this table below, taken from the paper by Hills, Thomas and Dimmesdale published in 2010, they break up the three century period into five sub periods.

In the most recent period, between 1952 and 1992, the average downturn lasted about three years, followed by an upturn also lasting three years.  That makes an an overall cycle of of just under six years.

Therefore if we do get a recession it may not be over quickly. In fact three years can feel like a very long time during a recession.

Why now?

If you read the Bank of England's inflation report for May 2019 the first graph in the report (below) shows this quite marked decrease in economic activity for both emerging markets and  advanced economies.

If we look at leading indicators it can give us a slight hint as to what will happen with growth in the near future. 

The Purchasing Managers indices (below) shows the manufacturing components have fallen below 50, which is a sign of contraction. But you can also see the composite, which combines manufacturing and services, is still above 50. Therefore the PMI indices just point towards a continued fall in the GDP.

When discussing the reasons for this fall the Bank of England and the Federal Reserve in the United States point the finger at trade tensions, in particular the trade war between the U.S. and China.

The direct effect of this has been on the bilateral goods trade between those two countries. The broader effect has been due to reduced global business confidence. The manufacturing sector has been particularly hard-hit. 

When companies are worried about economic policy and in particular trade policy they are less likely to invest. This results in slowed down economic growth but fortunately consumption growth has remained resilient.

In the U.S. the yield curve has been a fairly good predictor of economic growth. 

The Cleveland Federal Reserve Bank publishes a model which takes the difference in yield between short term bonds and long term bonds. As an output it forecasts GDP growth and the probability of a recession in the coming year.

This model is certainly not perfect as there have been two notable false positives. In late 1966 the US yield curve inverted but there was no recession and the same thing again happened in 1998. 

Below is the data that feeds into the model and the orange line is the steepness of the yield curve.

You can see that when it goes negative there is a grey bar that follows, usually within a year and that marks a U.S. recession. You can see in the graph the false positive in 1966 where the yield curve inverted and there was no recession. But on the whole when the yield curve inverts we normally get a recession within one year.

It happened in the 80s, the early 1990 and the early 2000s and then before the global financial crisis and you can see it's happening again now.

Now there's no economic theory behind yield curve observation,  it's just a relationship that people have noticed in the data.

Another problem is that there really aren't that many data points, there haven't been that many recessions since WW2. We only have about ten data points to work with, so the statistics aren't particularly robust. 

Below,  for what is worth,  the model currently forecasts a 40% chance of recession in the year to come. But it’s worth pointing out that the U.S. data is still okay. They still have very strong employment data, a fairly brisk wage growth and their PMI indicator is still above 50.

In the UK the National Institute of Economic and Social Research (NIESR) says that there is a significant risk that the economy is already in a recession that began in April and if we get a no-deal Brexit there is the clear possibility of a more material downturn.

Asset Returns

As an investor what we really worry about is what happens to asset returns during a recession and there is a fairly clear and consistent pattern with this.

On the x-axis in the top section of the graph below, I show the monthly return on U.S. Treasuries. You will see there are large positive returns of about 10% on the right and large negative returns of -10% on the left.

I have split the monthly returns into recession periods in red and periods of growth in blue.

What you can see is that during periods of growth, the total return on Treasuries is small and positive but during a recession the average return increases significantly. There is a shift to the right of this distribution and an increase in the size of the upside tail. 

You get more large positive returns during a recession and the reason for this is that people sell equities and buy Treasuries.

Treasuries are seen as much safer. They have a much lower volatility and they carry almost zero credit risk in the United States. We call that selling of equity and the buying of government bonds de-risking a portfolio.

n the bottom graph, If instead we look at equities the S&P 500  pattern is exactly the opposite during periods of growth.

The returns on the equity market tend to be stronger but during a recession the average return is negative.  If you look at the distribution of returns you can see that it's grown a large downside tail. In other words during a recession you're much more likely to have a large negative return and that drags down the average return for equity.

If we had a crystal ball that could predict recessions (which we don't) then the best strategy would be to de-risk just before recession hits and then just as the recession is ending switch back into equity.

In the graph below, we look at a broader range of indices.

As a I don't have such long time series for these, I've had to use a shorter time period, from 1995 to 2019.

I have ordered the assets from left to right according to the amount of the difference in return during periods of recession and growth. Recessions are in red and periods of growth are in blue. You need to look at the  black line which is the median return. 

For the Hang Seng Index you can see that during periods of growth the median return is positive, whereas during periods of U.S. recession the median return became negative.

The only one of these assets where the return turned positive during periods of recession was gold and that's because gold is also seen as a safe haven during periods of financial crisis.

I have covered whether gold is a good Investment in a previous blog and if you want to read more about it click here

If you want to actually buy those indices you can do so very cheaply through exchange-traded funds (ETFs). 

In the graph below, I've shown six of those ETFs from the iShares range, which is managed by Blackrock. These are the ones which had time series that extended before the global financial crisis.

What you can see is that UK gilts were the only fund which kept their value during periods of U.S. recession. The median is roughly the same during periods of recession and growth.

Whereas European Equity showed much larger volatility or variation in returns during periods of recession. The median return was also negative 

UK corporate bonds also sold off during periods of US recessions,  as did European Equity, the FTSE 100 and the S&P 500 tracker.

If we did have time series stretching back further for these funds, we would see would be a very clear split. 

As investors de-risk they would  move their money into safer funds, typically fixed income. The safest part of the fixed income universe would be government bonds, either issued by the UK or by the US.  But if you did want a safe version of a US Treasury you would probably have to go for the hedged version to get rid of the currency risk. This is because for US Treasury funds the volatility is so small that it's swamped by the currency volatility. 

The assets that would sell off would be the risky ones (in red) and that's the equity funds particularly the most risky which the emerging market stocks.

Strategy

What strategy should we adopt when we're thinking about recession?

Fixed Allocation

The fundamental problem is that we can't predict when recessions will happen. Of course we can look at PMI indices or we can listen to the central banks but unfortunately they don't know either.

There is no perfect model for recessions, so the simplest thing you can do is to create a diversified portfolio.

Find a risk level that you are happy with and use that to fix your equity/bond allocation. Keep your fees low, then simply drip-feed your money into your investments over your lifetime and completely ignore the news.

Although this may  sounds like a dumb strategy, for the majority of people in the past this has produced a good outcome.

The reason why you ignore the news is that over the very long term shares may give you a bumpy ride but if you simply keep your money invested, markets recover and erase those periods of large loss.

That is true for equities in red but also true of bonds in blue.  

Dynamic strategy - (vary allocation)

An alternative is to have a dynamic strategy where you vary your allocation according to what you think is about to happen in markets.

That is exactly what dynamic multi asset funds managers will be doing on your behalf.

With these multi asset funds you often pay a fairly high fee for expert fund managers to do that fiddling for you. They try and predict what's going to happen in markets and then they move your allocation accordingly. 

However, if we compare them with something like a Life Strategy 80% fund, which is in blue in the graph below, you can see that this Schroder dynamic planner portfolio actually under performed that fixed allocation even though the fee is considerably higher.

Many of these multi asset managed funds fail to beat the market and that's true of this Investec Cautious Managed fund which has also under performed The Vanguard Life Strategy 60%

Even one of the best performing funds like Artemis Income Fund is running roughly neck-and-neck with the cheaper and fairly dumb Life Strategy 80% fund.