QUANTITATIVE MODEL
Great Northern Capital Small Cap Growth, first offered in 1996, and Great Northern Capital SMID,
created in 2005, utilize proprietary computer modeling to identify companies experiencing positive
changes in their consensus earnings estimates. Empirical research links upward equity price movement
with positive change in consensus estimates. Small Cap Growth and SMID each provide an actively
managed portfolio of 50 stocks with attractive earnings characteristics, optimized to control risk relative
to the Russell 2000 and Russell 2500 Index, respectively. The logic based (non-black box) approach
features a time-tested alpha driver, unique proprietary risk factor work, and a strict quantitative,
objective, unemotional sell discipline.
The alpha driver contains a quantitative “Earnings Expectation” model that identifies equities with
favorable earnings estimate characteristics. The model performs a daily download of all publicly
available earnings estimate revisions and continuously analyzes trend of estimates, rate of change in
estimates, level of analysts’ agreement in estimates, statistical dispersion of estimates normalized by
historic dispersion, earnings surprises normalized by history and industry, and estimate breakouts.
Stocks are assigned a score between 1 and 99, with those in the 1-15 range creating a buy list,
typically around 250 in number. A quantitative “Universe Construction” model further identifies equities
with superior response to improving earnings estimates. Certain characteristics such as capitalization,
price volatility, and earnings reinvestment cause greater price response to estimate revisions in certain
stocks. The model will not purchase a stock unless the company has been followed by at least three
Wall Street analysts (not necessarily the same three) for a minimum of 5 years. A quantitative “Risk
Factor” model controls systematic portfolio risk relative to the Russell 2000 or Russell 2500 Index,
utilizing such factors as interest rates, inflation, investor sentiment, valuation, volatility, and Russell Index
sectors. The model selects a portfolio of 50 stocks, optimized to constrain overall portfolio risk factors
to within one standard deviation of the Index, in combination with the greatest alpha potential. When a
stock’s individual score deteriorates to 60 in the 1 to 99 proprietary scoring system, the stock is
automatically sold.
|
Great Northern Capital’s proprietary risk factor work is unique relative to industry standards and has
undergone significant enhancement during the past 5 years as a result of continual research, validation,
and assessment. Whereas risk factor analyses typically perform a 60-month regression analysis on the
historic performance of an index, the frequent rebalancing and turnover of the constituent names within
small-cap and SMID indices renders such analysis ineffective in evaluating risk. A high turnover index
today bears little resemblance to itself 12 months ago, 24 months ago, 36, 48, or 60 months ago,
which begs the question as to exactly what the typical analysis is controlling its risk against. Great
Northern Capital utilizes a dynamic proprietary model that takes the index holdings in real time, and
reconstructs those holdings back 60 months to perform the regression, resulting in a highly accurate
assessment of the portfolio characteristics versus benchmark risk. Additional analyses using the macro
factors cited above completes the risk factor portion of the model.
|
