The
Strategist System™ Tutorial
In this part of our Tutorial
we will start at the beginning, and then gradually lead the new user through
the more advanced aspects of our investment system. Other tutorials (under development) will describe the historical
evolution of our approach as well as special topics of interest, e.g., our
generalized candlesticks (RaDiSH Transform).
MTI (Market Timing
Indicator) and the “Maxwell-Boltzmann” thermodynamic analogy
As mentioned in the
historical sections of this tutorial, our approach to the stock market centers
on the use of a Market Timing Indicator (MTI) to measure the strength of
the market. Our MTI is actually
calibrated to QQQQ, rather than to the market as a whole. This is done for two reasons. First, QQQQ is one of the best ETFs (Exchange
Traded Funds) for relatively safe and profitable investing. It represents the 100 largest companies in
the NASDAQ, has a very high degree of liquidity (the difference between the ask
and bid prices is generally a penny), and is one of the easiest equities to
accurately model. Second, QQQQ is
highly representative of the ‘tech’ stocks, and the ‘techs’ are a very
important driving force in today’s stock market. Although the market can (and does) rise without the tech stocks
(e.g., the Dow can rise even while the NASDAQ languishes), we do not like to
invest in a market in which the ‘techs’ are weak – such a market will not
continue to rise for long.
The Market Timing Indicator
is calculated from parameters emitted by our Preprocessor (formerly a
complete Mechanical Trading System).
Because of the way in which it is derived, however, the MTI might just
as well stand for “Market Temperature Indicator”. The calculation is rather analogous to
obtaining the temperature of a gas by looking at the distributions of molecules
traveling within different speed ranges (Maxwell-Boltzmann
distribution). In our case, however, we
utilize stocks instead of gas molecules, and we use a spectrum of 10 primary,
and 100 secondary, portfolios (with very peculiar buy and sell rules), which
allow us to derive an estimate of market strength. Not only can we see how full each portfolio is, but we can also
look at the effective ‘curve’ that results when we ‘plot’ the population levels
as a function of portfolio ID.
The MTI value alone is of
little use; however, from it we can generate three related parameters that are
highly useful:
(1) Hedging
Vectors (#0, #1, and #2) and the ‘Prudent’ Hedging Vector
(2) Maximum
Long% Allocation
(3) Market
Color Code
The Hedging Vectors are
formulas for ‘asset allocation’, and they are directly derived from the
MTI value. HV#0 divides assets between
Long investments and cash, HV#1 divides assets between Long investments, Short
investments, and cash, and HV#2 divides assets aggressively between Long
investments and Short investments. The
Hedging Vectors become more aggressive as their # increases, and HV#2 is
obviously the most aggressive of the three.
There is also a fourth
Hedging Vector – the ‘Prudent’ Hedging Vector. This vector presents our most conservative asset allocation
recommendation.
This brings us to the MOST
IMPORTANT feature of our entire system – the Maximum Long% Allocation. This is actually a range of values, which
specifies the maximum percentage of capital that should be at risk at that
time. At the low end is the
recommendation for ‘prudent’ investors, and at the high end (the maximum) is
the recommendation for aggressive traders.
Generally, the user of our system should decide whether he/she should be
following the ‘prudent’ (conservative) allocation, the ‘aggressive’ allocation,
or, perhaps, happily in the middle.
Thus, if the Allocation is given at 10-45%, prudent investors should not
have more than 10% of their money at risk at that moment, and aggressive
traders should not have more than 45% at risk.
It is certainly permissible, however, to exceed these figures, but in
this case we would recommend that fairly tight stops be used so that if the
market plummets, and the stops are triggered, that the unprotected amount of
capital not exceed the Long% allocation guideline.
Our Market Color Code
is modeled after the Terror Alert colors used by the United States
government. Green signifies that
the market is relatively safe for investment.
Blue suggests a slightly weaker situation that may still be very
favorable for investment. Yellow
is a caution signal that should cause users to think about setting some
Stops. Orange is a ‘tricky’
color, and is often where strong movements can take place – either up or
down. Finally, the color RED
suggests ‘shorting’ or staying in cash.
The Market Color Code is derived from the Hedging Vectors, and the
Maximum Long% Allocation, and helps to make the overall market picture a bit
more vivid.
Our System offers three
different types of signals, depending on the time frame in question:
(1)
For QQQQ we offer a day-trading signal.
This signal is found in Table II of our daily Prediction Table (PREDTABL),
and it has the values: ?+, *, **, ***,****,?-,S, SS,SSS and SSSS. The **** signal is our strongest Long
signal, and SSSS is our strongest Short signal. The ***,**,* and SSS,SS,S signals are progressively weaker Long
and Short signals, respectively.
Finally, the ?+ and ?- signals are generated when our models sense an
extremely ‘flat’ market that lacks a clear direction. The accuracy of the ?+ and ?- signals is perhaps 55-65%, the
accuracy of the * and S signals is about 70% … based on back-testing over the past
4 years. The **,***,*** and SS,SSS,SSSS
accuracies are perhaps 80%, 90%, and 100%, respectively – again, based on
back-testing.
(2) Although
QQQQ is our only day-trading signal, we have swing trading
frequency signals for 1000 stocks and ETFs.
The average Long signal for these 1000 equities is about 11 trading
days, and the average Sell signal lasts about 17 days. These 1000 stocks (and ETFs) and their
custom timing models are known as our Grail System ™. These models are unique in that they utilize
the current MTI value in addition to technical information for the specific
stock (i.e., recent price and volume data).
Our Grail System signals have only two states: Buy and Sell. In general, however, Sell DOES NOT MEAN
‘Short’; the accuracies of our ‘buy’ signals are always quite a bit better than
the accuracies of our ‘sell’ signals.
(3) Finally,
we offer intermediate-term signals for about 1500 stocks and
ETFs. Our intermediate-term signals may
stay in the Buy state for weeks or months at a time. Unlike our Grail System (swing trading) signals, our
intermediate-term signals have 3 states: BUY, SELL, and HOLD. These are generally the best signals for
investors … as opposed to traders.
(1) Pick
your own stocks (preferably high PPF stocks) by looking at the ‘buy’ and
‘sell’ signals in our SIGNALS report (SIGNALS.TXT).
(2) Follow
one of our 30 official Portfolios that are contained in our PORTSTAT
Report.
(3) Follow
one of the 10 Preprocessor Portfolios that are shown in our NEWPICKS Report. These are not true ‘swing trading’
portfolios since they are generated by a rather unique method, but they can be
highly productive. Use of these
portfolios is recommended only for experienced traders who can perform their
own technical analyses. Many of these
stocks are thinly traded, highly volatile, and sometimes difficult to get in or
out of.
If you wish to use our intermediate-term
signals to buy stocks or ETFs, then all of these signals can be found in Section
2 of our MKTSTATS Report.
Although our intermediate-term signals have a very low trading
frequency, they will inevitably result in more drawdown – and so this approach
is only suitable for people with a true ‘investor’ mentality.
Finally, remember Key
Point #1 – keep an eye on the Maximum Long% Allocation figure!
Grail System
Stocks/ETFs (SIGNALS, PORTSTAT, PORTCORR)
Our Grail System™ has
complex, customized timing models for 1000 major stocks and ETFs, and
these signals (BUY/SELL) have average durations of about 11 trading days for
Buys, and 17 trading days for Sells.
The SIGNALS report shows all current signals, and signal changes,
for all stocks. In addition, the PPF
(Predicted Performance Factor) is shown for each stock, and this value
will help you select the best possible candidates if you wish to develop your
own Portfolios. Section 2 of the
SIGNALS Report shows the stocks broken down by Market Industry so that you can
diversify your portfolio to the extent that you desire.
In the PORTSTAT
Report, however, we present 30 different portfolios that try to automate
this portfolio construction process.
The portfolios are graduated according to the maximum number of
positions (stocks) that they can hold, and are divided into 3
progressively more conservative Classes:
(1) The
“H” Portfolios. These portfolios
result in the highest annualized gain, but do a lot of trading. Moreover, they will ‘short’ QQQQ to
varying degrees as a Hedge when the market is very weak. The drawdowns for these portfolios are the
highest of the three different Classes.
(2) The “non-H”
Portfolios are a bit less aggressive than are the “H” portfolios; they
will produce a lower annualized gain, will result in less trading activity, and
generally have lower drawdowns than do the “H” portfolios. Furthermore, they never do any shorting.
(3) The “C”
(Conservative) Portfolios are designed for the more prudent
trader, and are the most suited for use within retirement (IRS) accounts since
they perform the least amount of trading – and have the lowest drawdowns of
all.
Although the 30
official Portfolios attempt to adjust their number of positions in order to (at
least approximately) track the Maximum Long% Allocation, it is advisable to
give them some additional help. For
example, the use of margin should be severely limited when the MTI (or Max
Long% Allocation) is weak. Also, a few
protective Stops may be advisable depending upon the latest market Forecast.
If you are following one of
the larger Portfolios in the PORTSTAT Report (e.g., P15H, P20H, P15,
P20, etc). then the PORTCORR Report may be occasionally useful. The idea is to protect the large portfolios
by using available margin funds (e.g., day-trading margin) to make 3-4 quick
‘shorts’ so that the portfolio is effectively neutralized, that is, “hedged” by
having shorts that are likely to make as much money if the market declines as
the Longs are liable to lose. Such
short-term ‘hedging’ (1-day) may allow a large portfolio to be maintained
without excessive loss of capital on market ‘dips’.
Predicted Performance
Factor (PPF)
The PPF is a parameter,
which attempts to quantify the ‘quality’ of a stock, i.e., its probability of
making a nice gain without excessive risk.
PPF values are contained in the SIGNALS report (for our Grail
System signals), and in Section 2 of the MKTSTATS report (for our
intermediate-term signals). The PPF
value is, of course, only an ‘estimate’ of quality – and they can never
guarantee that the stock will in fact do well.
Still, when constructing your own Portfolio (whether from Grail System
or intermediate-term signals), it is best to pick stocks with the highest
possible PPF value.
Mutual Funds and the
Mutual Fund Timing Indicator (MFTI)
One of the original major
developmental areas in the evolution of our Strategist System™ was on the
construction of a Mutual Fund Timing Indicator. The MFTI value was ultimately derived
from our stock MTI, and this led to the creation of a Mutual Fund Report
(MUTFUNDS), which gave comparative performance rankings for over 400
different mutual funds, belonging to 15 different fund families. This Report is extremely useful for periodic
(e.g., quarterly) rebalancing of retirement accounts. We will gladly add more mutual funds and mutual fund families if
our subscribers wish to provide us with lists of what they would like to see.
The Market Status
Report (MKTSTATS)
The MKTSTATS report is
our master report since it contains ALL signals, plus a historical
(back-tested) summary of market strength (MTI) over the past 4 years.
Section 1 shows our (back-tested) MTI, Hedging Vector, Market
Color Code, and other pertinent parameters day-by-day since 8/16/02. This Section illustrates very well how our
timing algorithm works, and in particular shows timing signals for QQQQ for
each trading day. Remember, however,
that these are ‘optimized’ numbers … and not necessarily the actual signals
that we posted at that particular time.
Section 2 summarizes all of our intermediate-term
signals (and PPF values) for about 1600 stocks and ETFs. Only equities in the Buy or Hold state are
shown, however.
Section 3 summarizes all signals for ETFs (Exchange
Traded Funds) and includes both our Grail System and intermediate-term signals.
Section 4 summarizes all of our Grail System signals, and
thus effectively duplicates what is contained in our new SIGNALS report.
QQQQ
QQQQ is not only a very important, and popular ETF, but it
also has a prominent role in our timing system since our MTI value is ‘calibrated’
using QQQQ. Thus, our MTI value is
primarily a QQQQ timing signal. We
could have developed our MTI to measure the Dow, or the S&P, or even the
whole market, but we chose to focus on QQQQ because we prefer to invest when
the ‘techs’ are hot – and also because the market is more predictable when
viewed from a ‘tech stock’ standpoint.
Our QQQQ signal is thus
probably our most accurate signal.
Remember, however, that we actually have 3 different QQQQ signals:
(1) Day-Trading QQQQ signal.
This is to be found in Table II of our Prediction Table (PREDTABL).
(2) Grail
System™ QQQQ signal. This signal has an average Long duration of
11 trading days, and an average Sell duration of 17 trading days. This signal can be found in our daily
Forecast, the Market Status Report (MKTSTATS), and in our SIGNALS report. When we simply say that QQQQ is a ‘buy’ or a
‘sell’, then we are referring to this master signal.
(3) Intermediate-Term QQQQ signal.
This signal may go months without changing state, and being an
intermediate-term signal there are three possible states: BUY, SELL, and
HOLD. Intermediate-term signals can be
found in Section 2 of the MKTSTATS report, and (for ETFs) also in Section 3.
If one is
constructing long-term portfolios from our intermediate-term signals (Section
2 of MKTSTATS), then it doesn't matter as much whether or not QQQQ
is in a 'buy' or a 'sell' state. Instead, the important factor is keeping
the percentage of funds invested 'long' in rough accordance with either our
'prudent' or 'aggressive' Maximum Long% Allocation figures. Few of our
subscribers are utilizing our intermediate-term signals, however, and so
in the following discussion we will focus on strategies that utilize our
shorter-term (Grail System) signals.
Just as our
Market Status Report (MKTSTATS) is subdivided into sections depending
upon whether we have a 'buy' or a 'sell' signal for QQQQ (our Grail System
Signal), our 'best' investment methods have to be similarly 'tuned' to what
QQQQ is doing. Remember that our Market Timing Indicator (MTI) is
really a QQQQ timing indicator since that is the way that it is calculated in
the first place.
QQQQ 'buy'
periods:
===================
Method I-A
(our favorite approach for margin-enabled trading accounts):
(1) Utilize
'official' portfolios shown in the PORTSTAT Report -- or -- construct
your own portfolios by selecting high PPF 'buy' stocks from the SIGNALS
Report.
(2) Vary
the amount of money at risk in the portfolios by trying to stay somewhat
close to the Maximum Long% Allocation published each day in our Forecast
(3) Consider
doing some 'shorting' of QQQQ using unallocated funds whenever the Table II
signal is 'S', 'SS', 'SSS' or 'SSSS'.
(4) Utilize
the Table II (PREDTABL) signals as a guide to deciding which actions to
perform near the Open (e.g., 'sells'), and which actions to take later in the
day (e.g., 'buys'). For example, if we have an 'SS' signal then we would
want to sell at the Open and wait until later in the day to perform 'buys'
(possibly getting cheaper prices).
Method I-B
(alternate approach -- especially for smaller accounts and experienced
traders)
(1) Follow one of
the Preprocessor portfolios shown in the NEWPICKS Report.
Since some of these stocks are thinly traded, it is necessary to be careful
when getting in and out (e.g., using more 'limit' orders).
Method II (our
favorite approach for retirement accounts):
(1) Hold QQQQ
"long" the entire time.
(2) Use a 100%
allocation UNTIL the net rise in QQQQ has reached 5%. Then, drop
the allocation to 75% QQQQ and 25% Cash.
(3) Use a 75%
allocation UNTIL the net rise in QQQQ has reached 6%. Then, drop
the allocation to 50% QQQQ and 50% Cash.
(4) Use a 50%
allocation UNTIL the net rise in QQQQ has reached 7%. Then, drop
the allocation to 25% QQQQ and 75% Cash.
*** The new QLD ETF will allow more aggressive investors
to effectively use 'margin' within cash accounts like Rollover IRAs and other
retirement accounts.
QQQQ 'sell' periods:
====================
Method I (our
favorite method for margin-enabled trading accounts):
(1) Avoid
Portfolios entirely. Instead, short QQQQ the entire time UNLESS
Table II (PREDTABL) shows a "strong" positive signal
(**,***,****). In these cases cover the short near the Open (as soon as
QQQQ shows strength) and then re-short prior to the Close.
(2) Be
careful in using margin -- or leveraging 'short' ETFs like QID.
Method II (our
favorite approach for retirement accounts):
(1) Short
QQQQ with PSQ (a QQQQ 'shorting' ETF that provides no leverage)
(2) Use a 100%
allocation (or less) UNTIL the net QQQQ shorting gain has reached 4%.
Then, drop the allocation to 75% PSQ and 25% Cash.
(3) Use a 75%
allocation UNTIL the net QQQQ shorting gain has reached 5%. Then,
drop the allocation to 50% PSQ and 50% Cash.
(4) Use a 50%
allocation UNTIL the net QQQQ shorting gain has reached 6%. Then,
drop the allocation to 25% PSQ and 75% Cash.
*** The new QID ETF will allow aggressive investors to
effectively use 'leverage' to short QQQQ within cash accounts like Rollover
IRAs and other retirement accounts.
Note: Shorting QQQQ in retirement accounts (whether
via PSQ or QID) is a risky procedure and is NOT APPROPRIATE for most
investors. Instead, trading activity should probably be limited to
holding QQQQ Long during 'buy' periods. The use of the QLD ETF 'might' be
appropriate in some cases to further magnify gains during 'buy' periods.
Then, a cash position should be maintained during QQQQ 'sell' periods.