Markets on the Precipice - likely outcome?


In the aftermath of the violent selling yesterday on the JSE, which took the index down by around -3%, the largest single day selloff since the lows set in August 2015 mini crash, resulting in the market approaching critical support zones, many market participants will be earnestly wondering to themselves if this is the start of more to come. Whilst we can't predict the markets future movements, we can examine the current market action in relation to the past and draw some likely scenarios. Let's see what we can uncover.

Charting the Market

Although it's difficult to quantify many of the traditional technical analysis methods (patterns, trendlines etc.) due to their subjectivity, I still find them useful to apply. In fact, over the years I've come to rely more and more on subjective approaches; I've found that supplementing quantified data with subjective overlays can be immensely useful. So let's take a look at one of my favourite methodologies that has an uncanny ability to call market reversal zones – Fibonacci Retracements.

Fibonacci retracements are ratios from the Fibonnaci sequence that provide us with an indication of the markets likely reversal zones, or key support zones. Applying Fibonnaci to the JSE Top 40 index we find that the index is at the 23.6% Fibonacci critical support zone. And from this level, two things can happen: 1) support holds and price recovers or 2) support is violated and price starts its move down to the next key support level. To get a better understanding of the most likely outcome, let's dig deeper into our analysis.

What about Market Breadth?

One of my preferred breadth indicators, one I use extensively to guide risk management, is the percentage of stocks trading above/below their 200 day moving average. This is a powerful way to very quickly gain a clear and unbiased perspective of the market. So let's take a look.

We currently have a strong majority, around 77%, of JSE listed equites trading below their 200 day moving average. That means close to eight out of ten JSE equities are in bear trends. In other words, the current selloff is broad based - we're seeing confirmation across the board.

Despite the fact that we've seen worse readings in this indicator, for instance in 2008, there is a strong tendency for the market to rally at around these levels, as emphasised by the blue horizontal trendline. However, the possibility of further downside still remains.


Volatility Index (VIX) Unfazed

What's interesting to me is that the volatility index (as measured by the VIX in the USA) has been relatively impartial to the current weakness in the USA. Considering the intraday move in the S&P 500 yesterday, we would reasonably expect to see a greater upside spike in the VIX. But that didn't happen. So what's going on here? Well in my mind, market participants appear to be expecting a market bounce from these levels – the VIX is not factoring in the possibility of a market crash or severe downside moves.

And finally, our powerful Emotional Barometer

If you don't know what this is, then take some time to read up about it here. I developed this indicator some time back and have been successfully applying it to my trading ever since. In fact, Trading Stocks has been built with the Emotional Barometer as the key decision driver and we've enjoyed exceptional success using it: the daily battle plan portfolio that I trade live for clients is compounding at +1000% per annum and has a monthly winning rate of 100%! You can view my live performance here.

The Emotional Barometer reading for the JSE Top 40 index is currently 10, and historically we witness high probability reversals with readings at this level. Further, a positive divergence exists between the indicator and the index, and, the divergence is apparent in other oscillators that I track as well. Therefore, in the short-term at least, there is a strong probability that the market will rise.

Likely Outcome

So let's assemble our data and make some calculated calls:

1) The current Fibonacci support zone has held a number of times in the past and is thus quite significant.
2) Our market breadth indicator is likewise running into support that has historically held quite well.
2) The volatility index (VIX) is not factoring in a risk of a market crash.
4) In the short-term, our Emotional Barometer reading signals that we are likely in a reversal zone. We further observe a positive divergences across our Emotional Barometer and other oscillators.

Based on these data, the probability of a strong market rally in the short-term is high. However, once the oversold condition relieves itself, there is the distinct possibility that we will retest and even break through the current support level. Therefore, the likely rally should be viewed suspiciously and traded with caution.

Untiil next time, happy trading. 

    CAGR Confidence Intervals and Consistency

    Last year Rowan spoke about the importance of consistency in portfolio construction and today I'd like to expand on the concept a little. There are a number of reasons why consistency is important: 1) our performance pegged pricing structure is based on consistency 2) it's psychologically easier to trade consistent portfolios 3) consistent portfolios enjoy quick recoveries from drawdown and lower maximum drawdown's and 4) you're more likely to achieve the backtested annual return in any given year for a consistent portfolio. In this post I'm going to discuss how we measure consistency, analyse real client performance alongside consistency and provide you with a simple formula to predict your annual return interval.

    First let's explore how we measure consistency in QuantLab. There are a number of ways to measure consistency in performance, we decided to use the Coefficient of Variation of monthly returns which we refer to as the CV ratio. The CV ratio is simply a strategy's volatility divided by its return:

    Monthly Standard Deviation / Average Monthly Return

    From the above formula it's plain to see that as risk, measured by the standard deviation, increases so too does the CV ratio. Therefore, lower CV ratios are preferable to higher CV ratios, or said another way the lower the CV ratio, the better your risk-return tradeoff. When deciding between two strategies with the same CAGR, it's always optimal to choose the one with the lower CV ratio.

    To emphasise the importance of the CV ratio in strategy selection, I'm going to share the live performance of five of our client's portfolios ranked by the CV ratio in the table below.

    Client CV Ratios

    By analysing the table you'll find that the portfolio with the highest CV ratio is associated with the lowest CAGR and the worst maximum drawdown. However, the backtested CAGR of this particular portfolio is very high, it's precisely 70.06%. So why then is this portfolio performing far below expectations? Well one of the reasons is the high CV ratio, or high volatility associated with the CAGR. High CV portfolios exhibit what I refer to as lumpy annual returns. In other words, some years deliver exceptional returns while others perform below average. This type of performance can be difficult to weather psychologically because one has high annual return expectations derived from backtesting that are often missed due to the high volatility in the return stream. 

    As mentioned above, portfolios with low CV ratios are more likely to achieve their backtested annual returns, or they display narrower return intervals. To demonstrate this I'm going to apply a simple statistical confidence interval to two strategies with similar CAGR's but different CV ratios. The rule is known as the 68-95-99.7 rule and it measures the percentage of observations that lie within a band of one, two or three standard deviations around the mean. Below I've included the two strategies and their respective statistics.

    Strategy Confidence Interval

    For the band width I included the 95% rule i.e. we can be confident that 95% of annual return values fall within two standard deviations around the mean. The 95% band is simply the CAGR +/- two times the standard deviation, and the width is simply two times the standard deviation. As expected, we can conclude from the table that the strategy with the highest CV ratio also has the widest return confidence interval. In other words, the performance of the strategy is more volatile with annual returns varying more widely. The strategy with the lower CV ratio has a narrower return confidence interval and thus is more likely to achieve the backtested CAGR in any given year.

    If you'd like to calculate the confidence band of your portfolio simply subtract two times the annual standard deviation from the CAGR for the lower bound and add two times the annual standard deviation to the CAGR for the upper bound. The computed range provides you with a 95% confidence interval of future returns. (As an aside, this is a crude calculation that provides a quick estimate. One should really use the arithmetic mean and confidence intervals assume normal distribution which may not be the case.)

    The CV ratio is a powerful way to build portfolios providing many desirable attributes, both in terms of performance and trader psychology. We believe that the vast majority of traders are best served trading portfolios that are optimsed for consistency. Quantlab makes it easy to build, test and trade consistent portfolios. 

      Engineering a Synthetic Volatility Index – Part 2

      In last week's post I discussed two basic requirements for our proposed synthetic index: 1) it must accept price as its sole input for its calculation and 2) it must exhibit a high correlation with the VIX when applied to the S&P 500. An indicator that satisfies both of these requirements could be effectively used to measure broad market volatility in any global index, regardless of whether or not the index in question has a corresponding volatility index. Because changes in volatility often precede changes in market regime, it can be of tremendous value to monitor the volatility of your favourite market.

      A good place to start any search related to volatility is with the Average True Range (ATR). The ATR is based on price alone, and thus fulfils our first criteria. Developed by Welles Wilder, it measures volatility in price by taking an n-period exponential moving average of the true range. The true range accounts for gaps in its calculation by comparing moves form the prior days close. Its formula follows:

      Max[ (high – low), abs(high –prevClose), abs(low – prevClose) ]

      An index experiencing a high level of volatility will have a higher ATR relative to its past and vice versa. One of the downfalls of the ATR is that it can't be used for cross-sectional analysis: we cannot use the ATR to rank indexes from high to low volatility because an index with a high ATR may just be trading at a higher price. A simple solution to this problem is to divide the ATR by its closing price. This in effect normalises the ATR, giving us a value that can be used to compare the volatilities of a basket of indexes/stocks.

      My preferred method of measuring volatility is slightly different to Wilders to allow for the shortcomings mentioned above. It involves taking an n-period simple moving average of the true range divided by the close, or:

      Simple Moving Average (trueRange / close)

      This normalises the daily volatility and then averages it over our desired timeframe. The normalised value is returned in the form of a percentage: it is the average daily percentage movement in the index over our chosen timeframe. Because it is a normalised value - unlike Wilder's ATR - we can use it for cross-sectional analyses such as ranking from high to low volatility.

      Next I had to decide on a lookback period to employ. This couldn't just be any value, but one that when applied to the S&P 500 achieved the highest correlation with the VIX. After running some optimisation code I finally came up with a suitable candidate that has a correlation of +0.92. Below I share my results:

      We find that as we approach a parameter value of 20 for our lookback the correlation improves, while the correlation deteriorates for values above 20. Our optimal value then is 20, or:

      Simple 20 day moving average of (true range / close)

      Taking into account the sample size (I ran this test from 1990 to the present) and parameter robustness, I believe these findings to be statistically significant. In other words, we can apply our synthetic indicator to any market index and be confident that we capture the inherent volatility in much the same way that the VIX captures volatility in the S&P 500. The image below nicely illustrates the high correlation of our Synthetic Volatility Index with the VIX.

      For those of you that use Metastock, here's the formula to calculate our synthetic index: Mov(ATR(1)/C,20,S)

      As always we welcome your comments. Until next week, happy trading. 

        QuantLab Launch

        We have finally completed integrating QuantLab's execution capabilites and are set to have the official launch on the 26th March 2014 at the Durban Country Club. The seminar will start at 6pm and run for approximatley 1 hour where I'll be discussing a broad range of topics relevant to quantitative analysis, and in particular QuantLab. If you'd like to attend this seminar please don't hesitate to contact us.