The "Big Dipper Trough Locator" (BDTL) was a project to extract only the highest confidence JSE trough reversal signals, so that they could be used in conjunction with the other very high confidence JSE entry risk management systems such as BITS, STORM and QUANTUM3.

Each of these signalling systems approaches the JSE entry problem from different angles. No single high confidence signalling system is perfect or has a high frequency or exploits all the possible low risk JSE entry points, so building a suite of systems with which we can act confidently is the only route to go. BDTL has a low frequency, but it also has a low overlap with the other signalling systems which means its addition provides for more robust signal opportunities for our subscribers.

Apart from providing high-confidence JSE entry signals for investors, we will also show how BDTL can be used to run the most powerful medium-term trading system ever developed by us, yielding in excess of 1,500% return over 12 years with just 15 trades, as opposed to a JSE buy-and-hold strategy that would have only delivered 354% return!

1. THEORY OF OPERATION
The Big Dipper is based purely on daily JSE ALSH volume breadth readings. On a day-by-day basis, the following stringent quantitative conditions need to be met before it fires a BUY signal:

1. The UP/DOWN volume ratio for the day must meet a certain threshold "a"
2. The average UP/DOWN volume ratio for the last "x" days must exceed a certain threshold "b"
3. The average DOWN/UP volume ratio for the last "y" days must exceed a certain threshold "c"
4. The maximum DOWN/UP volume ratio witnessed in any day in the last "y" days must exceed "d"

The exact values used for a, b, c, d and the durations of x and y are proprietary to PowerStocks Research. All the above conditions together are excellent at identifying major troughs and the commencement of new bull markets.

If you think about it, these conditions ensure that with "a" we can be certain that a significant amount of buying entered the market on the day, and that through "b" that such buying is not a once-off affair but the culmination of a steady but robust growth in the number of buyers returning to the market after a sustained period of heavy selling pressure measured by "c" that included at least one major panic sell-off within the last "y" days as measured by "d" that flushed the last of the sellers from the system.

What does this leave us with? Think capitulation, disgusted sellers, lack of any more sellers and rock bottom prices. Think "smart money" rushing to snap up quality stocks at bargain basement prices. Think trough. Think new bull market. Think "scary roller-coaster down the hill and rapid reversal as it shoots up again". Think "Big Dipper time!"

2. SIGNAL HISTORY
The Big Dipper only found 25 occasions in the last 12 years where all these conditions were met. That is roughly two times per year or once every 6 months. They are shown in the signal chart below. The thick signal-lines are two signals at least 5 trading days apart and the vertical grid lines are each 30 trading days apart. Note there is a signal on the extreme left of the chart:



Whilst the Big Dipper does not detect ALL major troughs, it certainly gets MOST of them, and detects the start of ALL major and/or secular bull markets.
But more importantly, it does not get fooled by BEAR TRAPS. Look at the last six major corrections - each of them, about midway down, found support in a "bear trap", sucking in the buyers, before crashing further. This sucking in of unwary buyers that think the bear market is over is an important step in creating final capitulation and disgust required to flush all the sellers from the system and spawn a new bull market. In all of the bear traps on the above chart, there is not a single Big Dipper signal in sight!

An exploded view on more recent times, as the Big Dipper is shown in the PRO TIMERS section of the JBAR report, appears below:



You will note that the Big Dipper recently fired a signal, hinting at a new bull market phase of the more recent correction. The power of combined signalling systems is demonstrated aptly here since none of the other high confidence systems, as at this date are flagging anything at this time.

Also, of the twelve signals represented on the above chart, only the 1st, 3rd, and 6th signals starting from the left coincide with our other high confidence trough detector that measures selling intensity, STORM5+. In fact STORM5+ is very infrequent (but superb confidence) and only showed 4 signals in the entire period above:



Note that the Great Trough, in March 2008, deemed by many to be the real bottom of the Great 2008 Crash was detected by both the Big Dipper and STORM5+ with STORM5+ being a little premature and Big Dipper obviously being ever so slightly late. At PowerStocks, our view is that the trough of 20th November 2008 was the "lowest point" and only the Big Dipper detected that. Both systems saw the other huge trough just before this, in October 2008. Big dipper also fired a signal just before the October trough, which would have been devastating to a short term trader, but to a medium to long term investor THESE WERE ALL MAGNIFICENT SIGNALS!

3. ACTUARIAL PERFORMANCE
It is time for the numbers. Just how successful is the Big Dipper? It looks great visually on the graph of the ALSH index for the last 12 years but how did this translate into win-rates, average returns and gain/loss ratios for various holding periods? The tables below show The Big Dipper strategy together with three other very robust systems, namely STORM5+, STORM1 and QUANTUM3.



We note that the Big Dipper is not as kind as the other systems for trading periods less than 20 days, in terms of gain/loss ratios and average gains. The supreme winners for these short trades is undoubtedly STORM5+ followed by our QUANTUM3 flagship. For 20 days, Big Dipper is comparable but STORM5+ and QUANTUM3 again dominate the gain/loss ratio race from 30-50 day holding periods.

But it is after 50 days that the Big Dipper comes into its own, completely swamping the gain/loss ratios of any other strategy even though win rates are similar. This is indicative of the strategies' ability to pick out entry points that have sustainable medium to long term gains - its ability to detect the commencement of new medium term and secular bull markets after a correction.

4. BUILDING A POWERFUL TRADING SYSTEM WITH BIG DIPPER
Long-term investors will love the Big Dipper to load up their portfolios on the troughs. However, medium-term traders can implement a phenomenal trading system with BDTL- the most powerful ever witnessed by us in fact. It works by trading the JSE ALSH index through a SATRIX40 or other vanilla index ETF, or to really crank up the gains you can even use it with SATRIX40 Single Stock Futures (SSF's) or Contracts for Difference (CFD). It works like this:

1. Enter the  JSE the very next day after you see a BUY signal.
2. Ignore other BDTL signals while you are vested in the JSE
3. Hold for 80 trading days (not counting weekends) and then SELL
4. Stick your proceeds into a fixed deposit or call-account till next signal
5. Repeat process from (1).

This simple trading scheme would have resulted in 15 trades been executed since January 1998 from the total of 25 signals that BDTL generated. Since you effectively are only trading BDTL signals more than 90 days apart, that means we have a trading system that is likely to be different to the one with all 25 signals. The difference is shown below.

 

The actuarial chart on the right shows the trade statistics for the strategy that only looks at BDTL signals 80 days apart and ignores the rest. There are obviously less signals but the characterisitics of the new system is very different to the original. The new system is less forgiving than the old for trade periods less than 30 days which means you are going to have to stomach some volatility and even some losses in the early days of the trade. However over 60 days the new strategy really shines with its gain/loss ratio just like the original strategy.

The set of 15 signals that are at least 80 days apart and represented by the trade system in the right table above are shown below: Each trade length is exactly 80-days.



After exhaustively testing various holding periods in the BDTL timing strategy, 80 days delivered the best results over a period of 12 years in terms of win-rate, gain/loss ratios, total points gained (profits generated) and volatility/drawdowns. It is no surprise this holding period also holds the best gain/loss ratio below 100 days in the actuarial table.

Looking at the above chart we see that this trading system, whilst not a trend-trading system (like the wonderful Turtle Trend Trading system) certainly comes close to picking up all the major up-trends on the JSE. It certainly will latch early onto new bull trends that commence after a serious sell-off. The 80-day trade limit, whilst certainly responsible for the superior returns of this system does mean it bails out early from potentially longer up-moves. This is possibly the only criticism of this strategy. You can see above that it missed out on nice bull markets from 1999-2000 and mid 2003 to early 2004, the last three quarters of 2005 and 2006. Some of these were really long bullish periods it would have had us sitting on the sidelines. Fortunately, we advocate the concurrent use of multiple high confidence trading systems which means we would have picked these bull markets up with some other strategies.

The main difference between the Big Dipper and most Trend Trading systems is that Big Dipper gets in EARLY whereas most, if not all well known trend following systems get in much LATER. Also Big Dipper will be vested in the markets a lot less than the trend following systems. The reason Big Dipper manages to perform as well if not better than the best trend-following systems is that extraordinary returns normally accompany the first part of a bull market or recovery, a period when most trend following systems are still sitting on the sidelines waiting for new highs. In effect, Big Dipper enters on new lows (the lowest of lows in fact!) while trend system generally enter on new highs. This is why Big Dipper is an excellent companion with say the PowerStocks S2 Turtle Trend Trading System.

TRADE RETURNS PERFORMANCE
Let us examine the returns profile and volatility of this trading system.  Note we deemed 0.5% brokerage charges for the BUY and SELL transactions and the prime-rate of the day less 4% for interest earned whilst we were sitting in cash. To be realistic, we assumed JSE entries occured the day AFTER the trader would have seen us publish the signals.



The total returns are simply stunning, beating the SwissClock by 350% over the same period with 10 less trades. With this timing strategy, your accumulative time vested in the JSE would have amounted to only 39%, effectivley eliminating 60% of JSE risk. Your annual compound returns would have been 26.5%, almost double that of the JSE buy-and-hold strategy.

The chart on the right shows that trades, in the beginning are subject to some volatility though. Over 66 instances were observed where the 10-day rolling returns was more then 5% in the red. The worst drawdown we suffered over a 10-day period was a hefty 13.4%. To protect yourself from major corrections on a rare unlucky signal, you can thus set a rolling 10-day stop-loss of 10% on your trades without killing the strategy. Tests have shown that anything tighter than this literally halves the TRI (Total Return Index) of the system and is not adviseable.

The individual trades are shown below. Note the long waiting periods between trades (shown as days in the OUT column at the right of the table). Whilst many of these periods were brutal bear markets where we were quite happy to be sitting on the sidelines, some of them, as we discussed in the previous section, were nice bull markets where our 80-day limit had forced us out the JSE.



5. EFFECTS OF TAXES
Of course the active trading strategy with The Big Dipper is going to be subject to income tax as opposed to a ALSH buy-and-hold strategy only subjected to 10% capital gains tax when we finally dispose of the shares. We recomputed the TRI of the two strategies, but this time ensuring each strategy had to pay their associated taxes due at the end of each February.



We see that taxes are extremely debilitating to the the returns of the active trading strategy, robbing it each year of 40% of its profits that could otherwise have been used for compounding returns. The dip of each strategy at the right end of the chart above represents the taxes due should both portfolios have been liquidated into cash in February 2010.

Nevertheless, assuming straight trading in a vanilla ETF that tracked the ALSH on a 1:1 basis (no gearing or derivative trades), the after-tax returns of the trading strategy still comprehensively out-performed the ALSH buy and hold strategy. Remember, it did this by only being exposed to JSE risk 40% of the time and as you can see from the chart was in cash in most major corrections, meaning far less ulcers in the process.

The only time the buy-and-hold out-paced the trading strategy was during the biggest bull run in modern history from 2004 to 2008, where the JSE came from behind to take the lead from 2006 onwards. Alas, every bull market is subject to major recession borne crashes ansd after May 2008 the Big Dipper traders were smiling again.

6. CRANKING UP THE RETURNS
To overcome the effects of the tax and significantly boost returns of the trading strategy, it is advised to use ALSH futures contracts to make your bets. All trades are less than 90 days which suits the purchase of 3 or 6 month contracts perfectly.

As we have high confidence entry points and a 80% plus trade win rates, the trading strategy is quite suitable to leveraged instruments such as futures with 3, 5 and even 6 times gearing and we can effectively supercharge our returns. Due to the up to 10% drawdowns that can occur, we recommend going for lowest gearing possible. If you go with highly geared products we strongly suggest you run a stop-loss at twice the Average True Range (ATR) of the TOPI index at time of trade entry. Not only can most investors and traders not psychology cope with 50% drawdowns, but many geared products may terminate for loss of all capital below certain values (such as knockout warrants.)

We prefer the use of SATRIX-40 or SATRIX-RESI Single Stock Futures (SSF's) or Contracts for Difference (CFD's) or long-dated 4:1 geared TOPI warrants. These should all be available as one-click purchase shares with your online broker (just like any other share). CFD's are nice as you do not have to bother with the term left on the instrument (they can be held as long as you like) and offer lower gearing more suited to this strategy. There are CFD's available for SATRIX40.

Let's illustrate the use of gearing. With a SATRIX-40 SSF with 6x gearing, you only need to invest R10,000 to get exposure to R60,000 worth of SATRIX-40 shares. Let's assume with a Big Dipper trade, the ALSH rises 15% (the average winning trade over 12 years, with a 1.5 Sharpe ratio meaning it does not vary much). The SATRIX-40 will likely mirror 95% of the movement of the ALSH index (since its constituent shares dominate the JSE) which means your R60,000 will now be worth R69,000 for a R9,000 gain. Remove 40% of the gain to accrue for income taxes (although this is only due at the end of the tax year) and this leaves R5,400 net gain. This gain was achieved with a R10,000 investment which is a 54% after-tax return on investment!


7. REAL LIVE RECENT RETURNS
At PowerStocks Research, we run a live composite trading system called the ALSH PRO TRADER. This is simply three seperate hand-picked tradings systems from our arsenal combined together into one trading strategy execution. The three systems have been chosen for not only gains performance and reliability, but diversity in their underlying logic. This diversifies our trade risk but also offers a good degree of non-overlapping signals frequent enough to keep us trading at least once per month on average.

PRO TRADER is a perfect example of the PowerStocks philosophy of combining diverse, highly reliable and robust, but infrequently firing and non-overlapping trading systems into a composite strategy that offers superior risk reduction and returns. The systems currently part of the PRO TRADER suite are BIG-DIPPER, BITS-MT and QUANTUM3 (Bullish markers only). We are busy incorporating our new Turtle Trader S2 system to join them, but this is a lengthy process of testing.

With PRO-TRADER there is a pool of trading funds which are allocated to the various systems as they fire signals. Currently the allocation is simply 3 seperate but equal capital amounts dedicated to each system. Funds dedicated to one system are never allocated to another. PRO-TRADER-II offers better capital allocation schemes, since PRO-TRADER-I sometimes sits out the markets with a lot of cash, since it is only vested in the JSE 29% of the time, but this will take some time before we release it.

The chart below, which is published regularly in the JBAR PRO TIMER pages, depicts more recent performance of the seperate components of PRO TRADER since May 2007:


You can see how exceptionally powerful the returns are from Big-Dipper, having returned 108% since 31 May 2007 versus QUANTUM (bullish markers) which retruned 29.2% and BITS-MT which returned 46.5%. This is against the backdrop of the JSE only returning -0.8% over the same period! As a whole, when counting all 3 components of PRO TRADER together as a single composite trading system, PRO TRADER returned 61.2% versus the ALSH -0.8%, by only being invested for 30% of the total time!
 
Make a Free Website with Yola.