The Quantitative Solution to Beating the Benchmark
Market Structure EDGE™ was created by the team behind ModernIR.com. Founded in 2005 and today the largest provider of quantitative equity-market analytics to US-listed companies, ModernIR pioneered behavioral analytics for market intelligence. ModernIR is a nationally recognized market-structure expert. In 2017, the US House of Representatives Financial Services Committee requested founder Tim Quast’s recommendations for improving the stock market, which became part of the permanent congressional record.
Now we’ve created Market Structure EDGE™, a web-based application for incorporating market structure into portfolio planning, construction and shaping. The cornerstones of market intelligence are knowledge of the rules and seeing what the money is doing.
Where everyone else has focused on factors that identify better businesses, like earnings growth, or financial strength, we instead studied rules governing trade-executions and how prices are set. We built software, algorithms and mathematical models to measure behaviors behind price and volume – which proved predictive.
Market Structure Sentiment
Prices Behave Predictably Most Times
Core to our approach is a measure we call Market Structure Sentiment™ that uses proprietary behavioral data to predict when algorithms will stop setting higher prices or stop setting lower prices. It’s not mass psychology but how machines following rules calculate prices.
To beat the benchmark, you want to remove stocks that underperform it, which Overbought stocks do, and weight your portfolio toward what outperforms it, which Oversold stocks do.
Market Structure EDGE™
Think of it This Way
• Red: Overbought
Nobody will be 100% correct. But even half-right, a coin-flip, you could beat the benchmark by a wide margin.
When Broad Sentiment stops rising, the market is topped. We calibrated tops to 7.0/10, bottoms to 4.0/10. Data at left show 133 SPY tops, 154 bottoms, as measured by our Broad Market Sentiment (broad-market measures are less volatile than for a single stock, where 10/10 and 1/10 mark tops and bottoms). Shape portfolios with those. When there is neither, hold. Apply the data to the market or any stock.
*(SPY data Feb 1, 2012-Oct 15, 2018)
Market Structure Sentiment™ vs SPY
Proof We’re Correct?
First, we’ve been providing data-analytics to issuers for nearly 14 years. Market Structure Sentiment™ is a core predictive metric.
We’ve now applied our knowledge and experience to the market. With statistical computing language, we assessed significance in random groups of 40-100 stocks – the size of most portfolios – of Overbought and Oversold conditions.
Data tell the story. Stocks that are Overbought and become Oversold underperform the market by an average of 2.9%. Holding them hurts performance. Reduce weightings in Overbought stocks.
Stocks that are Oversold and become Overbought outperform the market by about 4.1% (Note: That the benchmark SPY is down during the analysis suggests buying Oversold stocks may be particularly advantageous in down markets – so weight your portfolio toward Oversold stocks to increase odds of outperformance.) Even when stocks don’t move from Overbought to Oversold or vice versa but instead fall along the scale, data show Overbought stocks have a propensity to underperform, Oversold stocks to outperform.
Length of Market Structure Cycles
Short but Sufficient Cycles
How quickly must you make changes to your portfolio? The mean cycle is 15 trading days. Trends form and fade rapidly in individual stocks yet cycles last long enough to permit shaping.
Stocks that are Oversold and become Overbought on average should be portfolio-shaped every month.
Load your portfolio into our platform and regularly track what’s Overbought and Oversold, and shift money from Overbought ones to Oversold components. If you have none, use our analytics to identify Oversold stocks to add to your portfolio – perhaps tilting toward the strongest sector. Enter at an uptick — say when a stock moves from 1.0 to 1.8. Exit on the 3rd 10 (three days at 10.0), or on the first downtick (when a holding moves from 8.0, to 7.6, for example). Wash, rinse, repeat. Most times this strategy will yield benchmark-beating performance.