You may have noticed a few new columns in the * Weekly Rewind*. This post explains each one for future reference.

I like the format that I’ve settled on because it provides a quick view of recent price action both in terms of broad market movement (through the graphs), and relative class behaviour (through the table).

It is important to stress the words “recent” and “relative,” because for many of the measures in the table, insufficient data points are provided for statistically confident results. However, by comparing the measures *across* securities and how they have change *through time*, you may come to find them quite helpful.

Regarding the headers, in most cases I have substituted more reader friendly names over true technical titles, as presented below. For time frames, 1-Week = Five Trading Days (Friday over Friday, unless it’s a holiday shortened week) and 4-Weeks = Twenty Trading Days. Where no mention is made, the default is Twenty Days.

**Performance Measure Descriptions**

**Price Change**– The easy one, Friday’s closing price divided by the respective prior period close.**Ten-Month Moving Average**– Research shows that semi-active portfolio management using secular trends is helpful in reducing draw-downs and volatility while prospectively enhancing returns and assisting in the identification of the strongest performing indices. These columns compare current price to the simple ten-month moving average, and then rank that difference across the tracked ETFs.**Travel Range**– This is the “Standard-Score” or “Z-Score“, which measures how far the last price is from the average over the period, as measured in terms of units of volatility (number of standard deviations). For traders, this is another way of seeing how close price is to standard Bollinger Bands (typically set at 2.0 SDs). Just remember that equity prices often exhibit “fat tails”, and rarely move according to standard statistical distributions.**Price Index (RSI)**– This is the same “Relative Strength Index“, or RSI, indicator shown in the charts for the SPY and QQQQ. (I didn’t want to confuse this unfortunately named measure with the others to its right.) This*price oscillator*moves between Zero and One-Hundred. The higher the measure, especially over time, the more overbought that security may be considered. Likewise, when the measure is particularly low over time, the more oversold.In Rewind tables, securities will be formatted

**green**(near oversold) when the RSIs are under 10 and 25 for the 2- and 5-day calculations, respectively. They will likewise be market**red**(near overbought) when they exceed 90 and 75, respectively.**Relative to S&P 500 (RS)**– This column compares the percentage price change of the SPY to the other ETF proxies over the respective periods. Traders often call this “True Relative Strength” or “Price Relative” to avoid confusion with RSI.**Correlation to S&P 500**– Correlation measures how closely the subject ETF moves either positively or inversely with the SPY on a scale of -100% to +100%. When the measure reaches 80%+, we can say that the two ETFs have a strong linear statistical relationship.**Relative Volatility**– This statistic, better known as the Beta Coefficient, measures the volatility of daily returns versus those of the S&P 500. It is calculated as the slope of the two return series plotted against each other. To the extent it remains stable over time, it can be thought of as a leverage ratio against the S&P.**Historic Volatility (HV)**– In this case the volatility measure scaled from 0% to 100% is what it says! The higher the number, the more volatile the ETF has been, both reflecting risk as well as often indicating a turning point in trend.**Trend Stability**– Moving further into the more esoteric measures, we have the “Hurst Exponent” estimation or inverse “Fractal Dimension.” At a high level, this statistic measures the level of “smoothness” or “persistence” of the recent directional trend (>50) versus a tendency to mean-revert (<50).**Risk Reward**– The relatively new “Omega” statistic goes beyond end-of-period risk/reward measures, such as the traditional “Sharpe” and related ratios, and effectively (albeit perhaps less intuitively )weights the distribution of returns over the subject period, giving preference to positive skew. If the result seem non-intuitive or wacky, consider (dis)confirming the measure by doing some mental math on the ratio of either “Trend Stability” or “Price Change” over “Historic Volatility.”**VIX % Stretch**– I occasionally refer to this statistic in my blogs, comparing the last VIX options volatility index to its fifteen-day moving average. Discussion of the usefulness of this typically mean-reverting indicator can be found throughout the Web. It is found in the lower right hand corner of the Rewind table.

As a reminder all the measures here are backward looking and therefore not necessarily predictive of future behaviour, particularly during inflexion points.

Discovered your chart today…Great stuff! I don’t fully understand what the “Risk/Reward” column says, though. Is a high number a good reward-to-risk ratio or vice versa?

A higher number indicates a more positively skewed return profile — a good thing. Because this particular chart uses only a twenty day window, it is best to compare the number for a given ETF relative to all other ETFs rather than looking for a set threshold. The ETF Rewind newletter provides the Omega and other Risk/Reward statistics for multiple time frames. Best, Jeff

Hi Jeff, great methodology! thanks for introducing such interesting statistics. May I ask you 2 quesions ?

- As far as I understood Omega ratio is a curve, how did you manage to sum it up with one figure ?

- How many observations did you use to estimate the Hurst dimension ?

Best, RaphaĆ«l

Hi Anon, good questions:

1) Imagine the curve you describe divided into a histogram with the bars above or below the threshold added and used to calculate a single ratio.

2) See my response above about time-frames. Normally you'd want a larger sample for statistical validity. However, the shorter-time frame may be instructive for rolling comparisons and intra-security analysis.

This is amazing stuff-very usable with time. HOW i WOULD USE IT IS ANOTHER STORY. Thanks for your stuff.

Frank Maris

Cleveland Ohio

frankmaris@gmail.com

Jeff,

Your subscription service requires Microsoft Excel. I have the spread sheet in Mac's iWork – Numbers 09. Will your spreadsheets open in Numbers 09?

LWS

LWS, I highly doubt it, but you are welcome to try a trial and report back. Best, Jeff

[...] (Click Image to Enlarge/ ETF Rewind Glossary) [...]

[...] (Click Image to Enlarge/ ETF Rewind Glossary) [...]

[...] (Click Image to Enlarge/ ETF Rewind Glossary) [...]

[...] (Click Image to Enlarge/ ETF Rewind Glossary) [...]

[...] (Click Image to Enlarge/ ETF Rewind Glossary) [...]

[...] (Click Image to Enlarge/ ETF Rewind Glossary) [...]

[...] (Click Image to Enlarge/ ETF Rewind Glossary) [...]

[...] (Click Image to Enlarge/ ETF Rewind Glossary) [...]

[...] (Click Image to Enlarge/ ETF Rewind Glossary) [...]

[...] (Click images to enlarge/ ETF Rewind Glossary) [...]

[...] (Click images to enlarge/ ETF Rewind Glossary) [...]

[...] (Click Image to Enlarge/ ETF Rewind Glossary) [...]