Market neutrality. Protected sectors – improving security indices.

by Insider Week team | Jun 15, 2020 | Trading articles

In the previous article, we talked about the “miracle of least correlations”, and how combining minimally correlated instruments greatly improves the risk/return of a portfolio. Today we will develop this approach further and talk about how to beat the S&P500 by combining the most stable and least correlated securities sectors.

We will find out which sectors are highly desirable in our portfolio and which are highly undesirable. We will measure the efficiency of investments through the eSharpe (annual rate of return/standard deviation) and CALMAR (annual rate of return/maximum drawdown) ratios, already familiar to us from the previous article.


To begin with, let’s consider what we should not keep in our portfolio so that we can invert the logic and come up with opposite, high-quality solutions later. Put simply, we should not take sectors with poor-quality yield curves. These are the sectors that earn less or the same amount per year as the S&P500, but with a larger standard deviation, greater account drawdown, and a correlation close to the S&P500.

An example of high-quality securities sectors compared to the S&P500:

Fig. 1: XLP (blue), XLU (red), XLV (orange), SPY (green)

First, pay attention to the annual yield (green box, Fig. 1), and then to the equally important parameter – the standard deviation (purple box). An extremely important point is that absolutely every sector generates a higher yield and a lower standard deviation than the S&P500. As a result, pay attention to the maximum drawdowns. All sectors in the list have lower drawdowns against the S&P500 with higher yields. Additionally, 2 out of 3 securities indices have a very low correlation to the S&P500.

Let’s calculate the eSharpe and CALMAR ratios to understand the annual return per risk unit (more details of these ratios you can find in the first article):
XLP – 6.00 / 12.15 = 0.493 eSharpe and 6.00 / 32.61 = 0.184 CALMAR
XLU – 6.74 / 14.82 = 0.454 eSharpe and 6.74 / 43.51 = 0.155 CALMAR
XLV – 8.03 / 14.16 = 0.567 eSharpe and 8.03 / 35.50 = 0.226 CALMAR
SPY – 6.03 / 14.92 = 0.404 eSharpe and 6.03 / 50.80 = 0.118 CALMAR

Each of the good sectors has better risk/return metrics than the S&P500. We will analyze the fundamentals of what is happening a bit later.

An example of low-quality securities sectors compared to the S&P500:

Fig. 2: XLF (blue), XLE (red), IYR (orange), SPY (green)

Let’s calculate the eSharpe and CALMAR ratios:
XLE – 3,02 / 23,97 = 0,125 eSharp и 3,02 / 63,91 = 0,047 CALMAR
XLF – 1,79 / 21,24 = 0,084 eSharp и 1,79 / 78,68 = 0,022 CALMAR
IYR – 8,14 / 20,63 = 0,394 eSharp и 8,14 / 69,68 = 0,116 CALMAR
SPY – 6,21 / 14,92 = 0,416 eSharp и 6,21 / 50,80 = 0,122 CALMAR

In this list, 2 out of 3 sectors have much lower returns than the S&P500 for the period (Fig. 2). But, more importantly, they have a much larger standard deviation and, as a result, higher maximum drawdowns in a crisis. Such sectors should be avoided as much as possible.


Why is it that some sectors behave systematically better than others? Let’s look at the fundamental features of each sector. We will start with the low-quality sectors so that we can better understand the high-quality ones.

ETF XLF – the financial sector, Fig. 3

Fig. 3: XLF (blue), SPY (red)

XLF comprises mainly banks. The problem with this sector is that it is strongly connected with credit money. During a recession in the economy, the percentage of loan defaults increases, and banks are forced to either restructure problem loans at a lower % rate reducing their profitability for current customers, or to deal with non-core assets obtained from the collateral of bankrupt customers. The influx of new clients is decreasing, because in a crisis, people do not tend to take a large number of new loans, and the diminishing number of new customers take loans at a reduced interest rate, which was lowered by the Central Bank to prevent a wave of bankruptcies in the crisis. In other words, banks are torn apart from 2 resonating sides: problems with loan returns from old customers and weak inflows plus reduced profitability of new customers. This is the reason why the banking sector has a high standard deviation of returns and very high maximum drawdowns per unit of earned returns. We should avoid such sectors as much as possible. The detailed composition of the XLF.

ETF IYR – REIT funds, Fig. 4

Fig. 4: IYR (blue), SPY (red)

Companies within IYR are engaged in the construction, purchase and subsequent leasing of commercial real estate. They are quite similar to banks. Companies that deal with mortgage real estate, are connected with a large number of loans, as a real estate unit is very expensive. Funds engaged in commercial real estate suffer on the one hand from a drop in the value of the real estate – their assets. On the other hand, they begin to lose clients in the face of tenants who in the crisis begin to leave the retail space rapidly, because they cannot pay a high rental rate as before. Here we are also dealing with resonating factors that sway the price. The detailed composition of IYR.

ETF XLE – the energy sector, Fig. 5

Fig. 5: XLE (blue), SPY (red)

XLE are mainly oil-producing companies, less often gas-producing companies. Prices of these companies’ securities have both the dynamics of securities in the market as a whole (they correlate with the S&P500), and the dynamics of prices for raw materials that these companies actively extract. And since any prolonged recession is a drop in world trade, the need to move a huge mass of goods from point “a” to point “b” decreases significantly. In other words, planes fly less, ships sail less, and trucks drive less. This leads to a drop in oil consumption and lower oil prices.

Fig. 6: An impulse correlation (violet) of oil (green) and S&P500 (black)

In a crisis, the securities of oil companies begin to resonate with the fall of the S&P500 and the fall in oil prices fig.6. This leads to the fact that we see such a terrible standard deviation and drawdowns in this sector. The detailed composition of XLE.

Now let’s look at the fundamentals of high-quality sectors.

ETF XLP – the wide consumer sector, Fig. 7

Fig. 7: XLP (blue), SPY (red)

XLP are companies that produce everything that can be found in the nearest supermarket, including the supermarket chain companies themselves. These companies are good primarily because they have a very weak drop in cash flow during the crisis. People are the last to save on food, household chemicals, and clothing. This all leads to the fact that XLP is one and a half times less sagging in the crisis and has a standard deviation much less than the S&P500, which makes it a great tool for your portfolio. The detailed composition of XLP.

ETF XLU – Public utility companies, Fig. 8

Fig. 8: XLU (blue), SPY (red)

XLU are those guys who are engaged in the supply of water, electricity, garbage collection etc. They provide the basic needs of people, and consequently, the volume of sales of their services hardly falls during the recession. In addition to higher returns with a lower standard deviation and a smaller drawdown, they also have the lowest correlation to the market among the ETFs discussed in this article. The detailed composition of XLU.

ETF XLV – healthcare companies, Fig. 9

Fig. 9: XLV (blue), SPY (red)

XLV – this is all extremely cynical: people get sick in any state of the economy, and this maintains stable sales of these companies’ services, regardless of the local state of the economy. Also, this sector has an excellent foundation from the fact that life expectancy is growing and the population is becoming richer over time. As a result, an ageing and wealthier population spend more and more money on health care every year. Excellent fundamentals and prospects. The detailed composition of XLV.

Now that we understand the fundamentals of high-quality and low-quality sectors, let’s put them together to look at the “miracle of least correlations” effect from the previous article.

Fig. 10: XLF+IYR+XLE [34%+33%+33%] (blue) vs SPY (red), monthly rebalancing

Although the index of combined low-quality companies has stopped losing to the S&P500 in terms of profitability (mainly due to IYR yield), it still has a large standard deviation and a greater maximum drawdown compared to the S&P500 (Fig. 10). The correlation of the portfolio to the S&P500 is very high. High standard deviation and correlation to the market do not allow you to fully realize the portfolio advantage.

And here are our high-quality sectors, also mixed in equal proportions:

Fig. 11: XLP+XLU+XLV [34%+33%+33%] (blue) vs SPY (red), monthly rebalancing

And here you can see the way the reduced deviation and correlation do their job (Fig. 11). We get a portfolio that beats the S&P500 at a standard deviation lower than that of the lowest volatility ETF on the list – XLP. The maximum drawdown of the portfolio is 1.5 times lower than that of the S&P500 and almost 2 times lower than that of the portfolio from the bad sectors.

Let’s compare the dynamics of two portfolios in comparison with the S&P500 on the same chart (Fig. 12):

Fig. 12: XLP+XLU+XLV [34%+33%+33%] (blue) vs XLF+XLE+IYR [34%+33%+33%] (red) vs SPY (orange), monthly rebalancing

Compare the eSharpe and CALMAR ratios:
XLF+IYR+XLE – 7.88 / 10.95 = 0.72 eSharpe and 7.88 / 33.56 = 0.23 CALMAR
XLP+XLU+XLV – 5.77 / 18.31 = 0.31 eSharpe and 5.77 / 63.89 = 0.09 CALMAR
As you can see, the low-volatility portfolio of their good sectors is more than 2 times better than the high-volatility portfolio of the bad sectors and it implements the “miracle of least correlations” much better (Fig. 13).


Let’s now combine our portfolio of good papers and treasuries using our knowledge from the previous article.

Fig. 13: XLP+XLU+XLV+TLT [19%+18%+18%+45%] (blue) vs SPY+TLT [50%+50%] (red) vs SPY (orange), monthly rebalancing

As you can see, we have the maximum drawdown of 1.5 times lower than that of SPY + TLT and 3.5 times lower than normal SPY, and at the same time, they did not yield to them in any way (Fig. 13)! Pay attention to the portfolio correlations. Our new portfolio correlates with the S&P500 2 times less than the SPY-TLT from the previous article. We have a virtually market-neutral portfolio that is sagging by less than 15% in the most devastating crisis of the century since the Great Depression!

In the next part, we will look at how to increase the profitability of our portfolio using “smart leverage”, protected against bankruptcy and paying interest rates for the loan to the broker.

All calculations are done in portfoliovisualizer. By clicking on the link, you can check all the results yourself:

See you soon and good luck in your business)

Hello Insider - our question is:
What do you think about this article?

We really appreciate your feedback.
Do you have questions or comments about the article? How do you implement the information from the article?
And do you want to find other such materials on
Please use the following form.
Thank you for your help!

very interestingless interestinguninteresting


YES - please also send me your newsletter about new materials.

I consent to the storage of my data.
Please prove you are human by selecting the Flag.