Help Center Thousands of today's Charts Free trial FT4Web installed on your computer Contact us: phone, address, e-mail Of Interest to Individual Investors Of interest to Money Managers Already have FT4Web Installed? Lots of well known FastTrackers Mutual Fund Charts
Google
Search the site for ticker symbols, fund families, stocks, or topics.
All the Web
Search FastTrack

Last edited 08/22/08

FTAlpha - FastTrack Alpha

FTAlpha is a selection method that combines return, risk, and correlation into a single indicator value.

  • FTAlpha encompasses the concepts of Modern Portfolio Theory, MPT.
  • FTAlpha deals with investing strategies including diversification and risk reduction.
  •  FTAlpha generally works best when selecting from a  universe of 40-50 mutual funds and exchange traded funds (ETFs), but will also work within  a portfolio holding dozens of large cap stocks from a universe of hundreds of stocks.
  • FTAlpha is usually used in Momentum Strategies. However, even when computing a Momentum Modeling, FTAlpha sometimes switch to "value" strategy when the low risk basis has a negative return for the period.

Click for a step by step example of actual use.

FTAlpha is implemented in two ways within FT4Web.

  1. Within the spreadsheet, the correlation portion of the calculation uses the low risk basis specified issue. Using the spreadsheet is how you actually do the modeling. See below.
  2. When using a Momentum or Value Model, the correlation is computed using the AVG being computed at that moment by the model. This calculation does not look into the future. Each day, the only the part of AVG that can be computed from the prior period's known historical returns is used.

Interpreting FTAlpha on the Spreadsheet Tab

FTAlpha will help you build a diversified portfolio of stocks or funds which performs well under volatile market conditions. The concept is simple: FTAlpha combines volatility, correlation, and return into a single indicator value. Assuming that your portfolio is adequately represented by the low risk basis line (see Mimicking below), then rankings of any family of funds or stocks using the FTAlpha column will separate those issues into three groups:

  1. Positive FTAlpha. These issues added to your portfolio should reduce risk and increase return. 
  2. Near zero value FTAlpha: Changing positions in these issues will not improve the risk/return of your portfolio except to the extent that you sell and use assets to purchase higher ranked issues.
  3. Negative FTAlpha: These issues when removed from your portfolio should reduce risk and improve returns.

Note: When the low risk basis line has a negative return for the period ranked, FTAlpha demotes  funds that have done well recently, and promotes funds which have performed poorly recent. Thus, even when using FTAlpha within a Momentum Model, when the low risk basis line has negative return, FTAlpha will cause the Momemtum Model to effectively use a Value Strategy. (Sell the winners, and buy the losers.)

Mimicking your portfolio

FT4Web does not give you a canned way to build an exact portfolio line, but "exact" is, generally, not necessary and undesirable.

  • If a single issue accounts for 40% or more of assets, just use that fund as the low risk basis (key in the ticker symbol)
  • If you hold 10 or more mutual funds, then likely your current portfolio resembles the S&P 500, Especially if most these funds are us equity funds growth, or value funds. Use VFINX as the low risk basis.
  • If you hold, for example,  5 funds with the portion of assets being, for example, 20%, 15%, 35%, 22%, 18%. Then put each fund in the ColorBar. Set the Appearance dialog settings to Monthly/None. Make the remaining cell, let's assume the blue cell, 'Screen AVG'. Make the low risk basis 'Blue Line'.
  • If most of your assets are in a few stocks, and you trade  regularly each week, then FTAlpha will not be of any value to you.

Starting the Momentum Modeling

While this page is about use of FTAlpha, the techniques described below for getting started with the Momentum Model can be applied to modeling by Return and Sharpe.

FTAlpha will help you build a diversified portfolio of stocks or funds which performs well under volatile market conditions. The concept is simple: FTAlpha combines volatility, correlation, and return into a single indicator value. For each period, FTAlpha computation is based on the AVG line that was computed by FT4Web's model .

  1. For each period, the lowest ranked FTAlpha issues are sold until 25% of assets are in cash.
  2. The assets are used to buy equal dollar amounts of the top 10% of issues based on the FTAlpha score. For example for a family of 40 issues, assets would be distributed among the top 4 issues.

Note: if you mimicked your portfolio as suggested above but actually the portfolio is,

  1. 60% in Fund X,
  2. and if the model suggests adding more Fund X,
  3. then, do not add more Fund X.
  4. Allocate assets to other high ranked funds.

In subsequent ranking periods, continue modifying the model's suggestions until your actual portfolio allocation is in line with the model.

Using the Spreadsheet to Duplicate the FTAlpha Momentum Model Results

At the end of the period when it is time to recompute the Model and make trades, DO NOT wait for the Model to tell you its trades. For example, if you are trading monthly, the evening of the 31st, after updating for the day, you would follow the spreadsheet instructions below, and place trades for execution on the close of the 1st. On the evening of the 1st, after getting the 1st's closing prices, then the Model will report the trades it made at the 1st's closing prices in the REBALANCE.LOG  . . . which should duplicate the trades you made at the close.

Assuming monthly trading, do the following on the spread sheet

  1. Load family that you used last month to make your trading decisions. Compute an FTAlpha Monthly Momentum Model AVG line for that family.
  2. Set the Low Risk Basis to the FTAlpha AVG line just computed. For example, if the FTAlpha AVG is the green line on the T Chart, then make the Low Risk Basis "Green Line"
  3. View the Charts. Set the poles for the period.
  4. Go to the spreadsheet and add the FTAlpha column.
  5. Right-click the spreadsheet and click 'Compute'. Click the header of the FTAlpha column to sort in order.
  6. Now you are ready to place trades with your broker. Remove or reduce the low ranked funds from your portfolio. Add the high ranked funds.

The Model trades 25% of assets, but you can vary that. If you trade too much of your portfolio, then you will lose the diversification which is essential. FTAlpha tells you what to add to complement your existing holdings, but if you trade 75% of assets each period, then there will be nothing left for the new holdings to complement.

You don't HAVE to use the same family every period. However, if you change the family, then you will not be able to exactly duplicate FTAlpha's results. If fact, we DO recommend that you change the family to add new interesting diverse funds . . . Don't change the family when trying to duplicate the Momentum model.

Weekly and Annual FTAlpha rankings rarely produce desired results. Monthly and Quarterly rankings work better.

Interpretation Spreadsheet Columns: Building a Portfolio

Display the FTAlpha column

  • Adding issues to your portfolio with high positive FTAlpha Values  will improve performance and decrease volatility  of your portfolio. 
  • Adding issues with positive FTAlpha and high correlation will improve portfolio performance, but with increased volatility. 
  • Adding Issues with Low FTAlpha will hurt portfolio return without reducing risk.
  • Adding Issues with Low FTAlpha and low Correlation will hurt return, but may substantially reduce risk. Bear funds often fall into this category.

Interpretation #2: Managing a Portfolio

In step 5 load a family of all your holdings. Add a Cor= column to the spreadsheet.

  • Increase your position in issues with high FTAlpha and low correlation. This will reduce volatility and increase return.
  • Increase your position in specific issues which have positive FTAlpha. This will increase return and reduce risk.
  • Increase your positions in issues with neutral  FTAlpha (hovering around zero) and low Cor=. Reduce  positions in issues with neutral  FTAlpha and high Cor=. This will reduce volatility without hurting return.
  • Sell issues with  negative FTAlpha and high Cor= .
  • Review issues with negative FTAlpha and low Cor=. These may be providing a comforting hedge (volatility reduction). Consider moving the assets to money market to achieve the same result.

LIMITATIONS:

This analysis presumes that the high/low correlation among family members will continue. Indeed, correlation is more consistent than return. Watch out for short term high flyers with low correlation and high FTAlpha . . . the low correlation will likely continue, but the high returns may fail.

FTAlpha depends on having a large number (30-50) diverse issues to choose from. Ideally, these would be issues that are about the same Standard Deviation( SD= ). If you add a money market fund to a small family of equity funds, within a few rebalancing periods, your portfolio will hold only the money market. Why? A Money Market fund has no risk and positive return. Thus the a 100% MM portfolio's return/risk ratio is very, very high and cannot be improved. Even if you start with a portfolio of 30 funds . . . there may only be 4-5 funds that exist on 9/1/88 and the portfolio will go 100% money market in the 1990 bear, and stay that way.

As with most technical analysis, funds are easier to analyze than stocks. Funds, including sector funds, will be more likely to continue their past volatility and correlation with respect other funds . . . stocks are much less likely to continue their trends. Adding a single, large fund position to your portfolio to reduce  volatility will likely be more effective than adding three smaller stock positions. However, in bear markets, it may be difficult to find funds, other than bear funds, which will help your portfolio.