How to Time S&P 500 Entries Using the VIX

Last Updated on January 26, 2023 by Mark Ursell

I recently received an email from Dr Alexander Elder in which he mentions using the VIX Volatility Index as an entry signal for the S&P 500. I particularly like that this trading idea is specific and measurable. In this article, I talk about the entry signal and show the results of testing this strategy.

Video

In this video, I talk you through the spreadsheet model I used and show you how I tested the different entries and filters.

Alexander Elder

Dr Elder is one of my trading heroes. His book: Come Into My Trading Room is one of my favourite trading books. I featured this book as part of my Essential Traders Library series of articles.

Trading Idea

“When the VIX is high, it’s safe to buy. When the VIX is low, go slow.”

The above is useful adage. It is similar to Warren Buffett’s famous saying: “Be fearful when others are greedy and greedy only when others are fearful.” We can use this advice to help guide our trading decisions, but it still leaves the questions: what is high, and what is low?

In his email, Dr Elder answers this: ‘VIX, the fear index, helps identify correction bottoms (i.e. buying opportunities) when it rallies above its upper Bollinger band.’

As soon as I read this, I realised that I could test this idea. I like the work of Dr Elder, but I will only trade a strategy that I can test myself.

How Can We Test this Idea?

“The devil is in the detail.”

The thing about most ‘trading rules’ is that they are impossible to either prove or disprove. Dr Elder has given us a trading idea based on a specific, measurable event (the VIX trading above its upper Bollinger Band).

We can look back through the historical data and set up a test for this idea. However, the problem that every trader and backtester comes up against is that an entry signal does not make a trading strategy. The detail is always important. We need to know things like position size, alternate entry rules, stop-loss and profit target distances, and other exit systems.

The way I deal with this is to minimise the other stuff to begin with. I want to try to find either an entry system or exit system that performs well without relying too much on clever trade management or position sizing systems.

How did I Test it?

I tested the VIX volatility entry using a Tradinformed Backtest Model. Tradinformed models are spreadsheet models built using Excel. They are a good option for you to test your strategies on different markets and different timeframes. If you want more information, check out the FAQ on How to Use a Model or see the latest models in the Tradinformed Shop.

What Settings Did I Use?

Entry

Dr Elder suggests buying when the VIX rallies above its upper Bollinger Band. So I decided to try two different variations:

• Break Above Upper Bollinger Band – Enter the market when the VIX breaks above the Bollinger Band. This is the most aggressive form of entry. This allows you to enter the market as soon as a potential move has started.
• Break Below Upper Bollinger Band – Enter the market when the VIX breaks below the Bollinger Band. This is a more conservative entry. This entry will get you in after the market is recovering.

Exit

I set up the spreadsheet to exit the position when the S&P 500 closed above its upper Bollinger Band. I did not use a stop-loss or profit target, but these can be added in later.

Filters

As well as trying two different entry systems, I also investigated two different filters. These were an experiment to see if the initial results could be improved.

Filter 1 – Linear Regression

The 200-period linear regression line provides a filter for the long-term direction of the market. I use the slope of the Linear Regression to avoid taking trades when the market is falling.

Filter 2 – Exponential Moving Average

The second filter is a short term EMA. The idea is to buy when the price is temporarily weak – you can set this to any value

What Results did I Get?

As shown in the video below I tested both entry types. The results were fairly similar, although there were some differences. For simplicity, I am only showing the results of the break above the Bollinger Band.

[table caption=”” width=”570″ colwidth=”160|150|130|130″ colalign=”left|center|center|center”] ,No Entry Filter,LR Slope,LR Slope & EMA
Gross Winning Trades,” $314,871 “,” $282,032 “,” $209,829 “
Gross Losing Trades,” $-160,474 “,” $-105,316 “,” $-35,298 “
Net Profit,” $154,397 “,” $176,716 “,” $174,531 “
Net % Gain,107.0%,113.9%,113.7%
Profit Factor,1.96,2.68,5.94
Winning Trades,63,53,33
Losing Trades,29,19,11
Win Percentage,68%,74%,75%
Ave Winning Trade,”4,998″,”5,321″,”6,358″
Ave Losing Trade,”-5,534″,”-5,543″,”-3,209″
Largest Winning Trade,” $17,972 “,” $19,549 “,” $19,854 “
Largest Losing Trade,” $-25,505 “,” $-11,845 “,” $-11,756 “
Max Drawdown,29.0%,22.0%,12.5%

[/table]

Conclusions

The strategy shows decent profitability and has a high percentage of winning trades. However, as shown in the video below, the strategy is prone to some large losses. These potential losses occur when the market falls and does not recover. Robust trade management would be needed to avoid the chance of catastrophic loss. The filters that I tested did help improve the profitability of the strategy and may be worth considering as part of a live trading strategy.