access to model readings and commentary
Header

Detailed Model Explanations

Trend Model

The Trend Model presents our position recommendations for the 180+ investable indexes in our universe.  For each index, we have separated the Short-Term score (top lines) and the Medium Term score (bottom lines).  Each of the indicators is composed of 20 distinct technical indicators with optimized parameters determined by econometric tests run in the R statistical software program. The Short-Term score is comprised of indicators parameterized for a 1 to 3 week horizon while the Medium-Term score uses indicators targeting a 1 to 3 month horizon.

Scores are summarized into five categories:

Buy (green boxes)
Accumulate (yellowish-green boxes)
Equal-Weight (yellow boxes)
Reduce (light red boxes)
Sell (dark red)

Buy signals are generated as the scores move from “Sell” or “Reduce” towards “Accumulate” or “Buy”. Similarly a sell signal is generated when the indicator moves in the other directing, passing from a green box towards a red box. An Equal-Weight reading recommends holding onto an existing long (short) position or waiting before entering into a new long (short) position. The yellow “Equal-Weight” reading does not trigger a buy signal nor a sell signal. In general, an investor will want to trade in-line with model signals.

The final two columns show the 21-day and 126-day highs and lows for the index. 21-days corresponds to one trading month and 126-days corresponds to 6 trading months. This allows the investor to situate the index price within the current cycle. These levels offer a rough idea of resistance levels (upside potential) and support levels (downside risk). A “0% upside” (or “0% downside”) would indicate that the index is breaking out to new highs (or breaking down to new lows), or “trending”. These ranges give a good initial approximation of the current risk/reward ratio for range-bound trading (non-trending).

Sector Allocation Model

The Sector Rotation Model applies a similar quantitative methodology as the Trend Model. In this case, the model attempts to scan a large universe of indexes and stocks in order to generate a list of candidates for inclusion in the portfolio (an “idea generator”). The Sector Rotation Model considers the risk environment and the relative trend strength indicators to calculate a new index price (“WMA index price”) which takes into account the strength of the asset relative to our dynamic, multi-asset composite benchmark (which includes the MSCI All-World index). We then co-integrate this WMA price vector with the index (or security) market price and run the two cointegrated vectors through our battery of tests. From this, we arrive at a score for each index or security. Scores are scaled into five categories:

Dark green = strong outperforming security
Light Green = outperforming security
Yellow = performing in-line with the market
Light red = underperforming security
Dark red = strong underperforming security

Again, the model separates Short-Term score (top lines), targeting the 1-3 week horizon, from the Medium Term score (bottom lines), targeting a 1 to 3 month horizon. Readings are given over the past five days to better detect the moment when an index/stock accelerates (decelerates) from the underlying market trend. As an index/stock moves from Under-Weight towards Over-Weight, the investor should consider the addition of the specific security to his portfolio. Conversely, as an index/stock moves from Over-Weight toward Under-Weight, the investor should consider dropping the security from his portfolio. This tool is useful in asset allocation in allowing the investor to clearly detect outperformance within a large universe of securities. The objective of the Sector Allocation Model is to create and maintain a fully-invested portfolio.

The final two columns offer another tool for rotating into and out of investable indexes. We use Relative Strength Indicators (RSIs) over 9-day and 14-day periods and Bollinger Bands to detect indexes that reach oversold (OS) and overbought (OB) level.  This analyze is conducted on both absolute index prices and our WMA price levels.  We recommend using the oversold and overbought readings on the WMA prices to serve as alerts on beaten down sectors (OS) and on sectors that may be over-heating (OB). The oversold and overbought readings on the market price may used in trading to lighten up (OB) on a position or find an attractive entry point (OS), from the long side.

Market Risk Indicators

The Market Risk Indicators attempt to detect changes in market participants attitude towards risk. A selection of market-based data is run in panel regressions with the variables returning the highest betas retained for inclusion in the indicator.  Two dependent variables — the NYSE Index and the DJ Stoxx 600 – are used to create a risk indicator for the U.S. market and a risk indicator for the European market. The selected variables are then re-run in an OLS regression to determine the weightings in the indicators. The variables are grouped into four clusters. First, we use binomial index pairs, with one member serving as a high beta or risk-on proxy and the other member a low-beta, risk-off proxy.  When the former strengthens relative to the latter, the pair contributes positively (greater appetite for risk) to the composite risk indicator. Conversely, when the low-beta sector strengthens relative to the high beta sector, the pair contributes negatively (lower appetite for risk) to the composite risk indicator.  The second cluster includes volatility indexes.  The third cluster is comprised of credit spreads. The final cluster groups exchange data.  The variable included are ( signifies variable in European risk model):

  • Nasdaq Composite vs. S&P 100
  • S&P Consumer Discretionary Index vs. S&P Consumer Staples Index
  • S&P Industrials vs. S&P Utilities
  • Russell 2000 vs. S&P 100
  • Nasdaq Biotech Index vs. S&P Healthcare Index
  • Philadelphia Semiconductor Index + DJ Internet Index vs. DJ U.S. Telecom Index
  • S&P 600 Small Cap Energy Index vs. AMEX Energy Index
  • IBOX High Yield Bond Index vs. U.S. 5-Year Government Bond contract
  • Stoxx Cyclical Index vs. Stoxx Defensives Index
  • Stoxx Industrials + Stoxx Chemicals vs. Stoxx Food & Beverage
  • Stoxx Banks + Stoxx Automobiles vs. Stoxx Healthcare
  • DJ UBS Industrial Metals Index vs. S&P Goldman Sachs Precious Metals Index
  • Australian Dollar vs. Japanese Yen

—————-

  • VIX volatility index (S&P options volatility)
  • VXN volatility index (Nasdaq options volatility)
  • RVX volatility index (Russell 2000 options volatility)
  • VXV volatility index (VIX 3-month volatility)
  • V2X volatility index (Euro Stoxx 50 options volatility)
  • V1X volatility index (Dax options volatility)
  • VCAC volatility index (CAC40 options volatility)
  • VFTSE volatility index (FTSE 100 options volatility)

—————-

  • CBOT Equity Put/Call Ratio
  • CBOT Total Put/Call Ratio
  • NYSE New 52-weeks highs
  • S&P 500 Advance/Decline line
  • NYSE Advance/Decline line
  • EuroStoxx 50 Advance/Decline line
  • S&P Europe 350 Advance/Decline line
  • Stoxx 600 Advance/Decline line
  • Percent of NYSE stocks about 200-day moving average

—————-

  • Corporate yields on BBA bonds less U.S. Government 10-year yields
  • Corporate yields on BBA bonds less Investment Grade AAA yields
  • TED spread
  • PIGS government bond yields less German Bund yields
  • European Corporate bonds yields less German Bund yields

The above indicators are dynamically-weighted to produce a score ranging between 0 (most risk averse) to 100 (highest risk appetite).  The direction of change of the indicator is of equal importance to the absolute level.  A rising index suggests a greater appetite for risk while a falling index implies greater risk aversion.  The chart may also be used to detect any divergences between the risk indicator and the broad stock index. We recommend accumulating / adding new positions only when the indicator is rising, while holding onto positions as long as the indicator remains solidly in “Risk-On” mode. A reading above 65 confirms a fully-invested portfolio. Based on portfolio risk tolerance, readings between 35 and 65 may justify lightening up on risky assets. Readings under 35 signal investor capitulation and a excellent time to deploy cash.

Sentiment Indicator

Our sentiment indictor uses a mix of market-based readings and investor sentiments surveys. The components and weightings are determined through a panel regression. Among the sentiment indicator components, we have included:

  • VIX volatility index
  • VNX volatility index
  • CBOE  Equity Put/Call Ratio
  • CBOE Composite Put/Call Ratio
  • McClellan indicator levels on S&P 500, Nasdaq-100, and Russell 2000
  • American Association of Individual Investors Sentiment Readings
  • Market Vane Survey Data
  • Investors Intelligence Investor Sentiment Data

We have delimited an “Extreme Optimism” zone and “Extreme Pessimism” zone. We consider a rising sentiment indicator reading as bullish. However, when the indicator enters into the “Extreme Optimism” zone, we seek to lighten up on long positions during rallies.  Conversely, we consider a falling sentiment indicator reading as bearish. However, when the indicator enters into the “Extreme Pessimism” zone, we begin to look for buying opportunities.

As a final note, the Market Timing and Sector Rotation models may be customized for a client with a unique benchmark (national or specific sector). The models can also be repopulated with any universe of stocks. Contact us to discuss customization.