Technically Speaking, February 2012

The Dow has had quite a rollercoaster ride ever since the previous all-time high of 11,750.28, back in January of 2000. This January marks the 12 year anniversary, which is 12 squared months (144 months), an observation that will come in handy a few charts from now. What a great time for a large-scale geometric perspective.

Can we fit the market into a geometric shape? And where and how would we start? How about at the very beginning of this big mess: the initial down trend from the previously mentioned high in Jan ’00 to the major low of 7197.49 in Oct ’02, expressed as a vector from high to low. All predominant reversals since are geometrically related to this vector, as we shall soon see.

The first chart shows the market conforming well to a large pentagon drawn from this vector (for a detailed breakdown of the actual technique of drawing this, skip to the end). The low of Mar ’09 is captured by the pentagon’s external circle. The time of the general support area of June ’10 (point 4) is indicated by the lower right corner. The upper right edge provides general resistance, although price is finally poking its head out. This helps to identify that current price action is at a very crucial point. A real break-out of this shape could be interpreted as a strong bull market with the above external circle and 45 angle providing potential resistance, while a reversal here could establish an important rally top.

The top of last May and current price are not just at pentagonal resistance, but directly at specific angles from the upper right corner of the pentagons’ surrounding rectangle. Notice that 36 degrees is a pentagonal angle (the upper sides of the pentagon are also 36 degrees from horizontal), and thusly, current price is at a harmonic AND thematic location.

In addition, by turning the internal pentagonal angle of 108 degrees into a time count in months, we arrive at points A (Jan ’09) and B (Jul ’10), both 108 months from the vector high and low, respectively. Both points are exact reversals: more support for a pentagonal approach.

The second chart shows the shear elegant power of the 45 degree scaling ratio, a true 1×1 relationship, as dictated by the 72 degree vector ratio. It reveals all important highs and lows as being connected via 45 degrees, but again, only shown in this particular fashion by aligning the initial vector at 72 degrees. Aligning the vector to other important degrees/shapes, such as 60 (hexagonal) or 45 (square), reveals important relationships as well, but I digress…..

If the Dow successfully breaks the pentagon in the first chart, I would use the ascending 45 degree channel for S/R. General resistance for Feb-Apr is 13,300 – 13,400.

The third chart is a necessary precursor to the fourth chart that clearly shows the relationship between actual time movement and percentage retracement levels of the pentagon height. Points 4 & 5: the vector’s 33 months, along with the 77 months between the vector low and the Mar ’09 bottom, identify the low and high of the previous up trend from Jul ’10 to May ’11 when expressed as retracement levels. These reversals are also harmonic with other major turning points using additional multiples of 11, at points 3 & 6.

The uses of trigonometric functions (ratios of triangular proportions) of pentagonal angles can work with alarming accuracy. At points 1 & 2, the sine’s of 72 and 54 (inverse of 36) degrees work well, identifying the all-time high with the main angle 72’s sine (.9511), and an important rally top at the sine of 54 (.809) which crashed the market (incidentally, 809 is 5 x 1.618).

All of these relationships clearly indicate that we are on the right track. But is there a harmonic explanation for current price? And what about timing points, since price techniques can be used for time as well?

The fourth chart reveals that current price is indeed at a junction of price and time. The fractional price level of ¾ is an interesting number: 12,722.72, (regarding 72 degrees), and is thematic with the other fractional levels of 1/10 and 1/3 (a.k.a 33%), providing accurate bottoms in Mar ’09 and Jul ’10 and even back as far as Oct 2004 (point 1). In addition, current price is at a fractional time point of 5/6. 3/4 price meets 5/6 time. This is a great location for a reversal, but certainly NOT a sure bet. Wait for confirmation, like a reversal candle pattern. If this resistance point does not hold, the market must have a lot of oomph!

This chart also reveals a ‘harmonic history’, regarding the 1/3 price level and its harmonic counterparts of 1/3 and 2/3 time points. At point 1, an important reversal formed exactly in price and time. At point 2, a potential top is foiled after the level is reached perfectly in time (2/3), then holds as support later on. Either of these two dynamics will play out currently. Just looking back in time….

Another clue to the power of the current time point of 5/6 is that it also aligns with 12 squared, 144 (remember?). In and of itself, this number has potential, as 144 is double 72, the pentagonal angle of the day, and both are naturally 2/5 and 1/5 of 360 degrees respectively. But when the squares of 4, 7, 10, and 11 also provide exact reversals, it has added weight.

Notice that the timing point of the sine of 33 degrees is 1 period from the all-time high. The sine of 33 will come in handy soon.

The fifth chart expands on the square numbers count from the previous chart. By geometrically relating the low of 11 squared on Feb ’10 with current price at 12 squared, they can be seen in a larger context of matching geometric relationships. The major highs at points 1 & 2 are perfectly connected to their major pullbacks (the pullback on Oct ’02, although not the lowest low, is very important, due to the extent of its retracement, and the almost vertical climb emanating from it) as these two lows vertically square with the vector high of 11,750.28, then proceed 45 degrees NE to the highs. Point 3 shows a smaller low (but impressive none-the-less) which gives the matching powerful rally-top of Mar ’08. This gives weight to a similar relationship forming at point 4, and supports the notion that the Dow is forming a secondary rally top, with a strong move down to follow. As noted before, a break of resistance at this crucial point would be very revealing.

Next, a 3 chart series demonstrating the power of the circle is presented (for great circle work, check out Michael Jenkins). These circles emanate from the initial vector at 72 degrees, and thus reveal relationships only seen from this relative perspective.

In the first chart of the series, the initial vector circle (blue) reveals harmonic vector distances to both the high of May ’06 (at 66 degrees), and the low of Jul ’10 (30 degrees). It is interesting to note the specific angles: 66 is double 33 and 30 is 1/3 (33) of 90 degrees. Secondly, the use of the sine of 33 (which gave us the all-time high area in the fourth chart) is used as an extension ratio, by adding it to 1. The all-time high did manage to poke its head out and look around, and so perhaps current price is just doing the same. The intersection of 45 degrees with the initial circle provided a bottom area on Feb ’09 (although the actual low was in early March, February had the lowest close). This indicates that the next intersection this July is a contender for a strong reversal. The second chart in this series displaces the circle by moving its center from the vector low to the major low of Mar ’09. It is a ‘graphic calculator’ of sorts, a measuring tool based on the initial vector. The top of Apr ’10 is this same vector distance, just at a different angle of ascent. This circle’s extensions of 1.33 (there’s ‘33’ again) and 1.618 (the Fib ratio is abundant in a pentagon) are telling. Notice how close the
actual price levels of the tops of these circles are to tops (in two cases, less than 3 points away). Once again, current price is hitting major resistance offered by the 1.33 circle and its top.

The third chart in this series is the gold (hey, speaking of gold, skip ahead to the last chart…). The initial vector circle (blue) has been displaced upwards to the first pullback (3-point low) in Jul ’09, immediately identifying the May ’10 top at 11 degrees. The up trend from the previous chart was at 7 degrees, and 7 + 11 = 18 which is the inverse of 72 (of course 7 & 11 are for a different paper altogether). Current price is hitting resistance at 16.5 degrees, which is a 1.5 ‘extension’ of 11 (11 x 1.5 = 16.5). Yet another harmonic reference to current price action.

If price successfully breaks resistance, the above red circle is the next resistance, provided by the next low in the up trend, on Jul ’10 (exactly 1 year after the previous low used). Successive measured moves from circles. 

The next chart shows no less than 7 measured moves (circular), and I even left out a few. In other words, the initial down trend (vector) has been repeating ever since it formed, and still is!

The next two charts give a weekly perspective using the massive downtrend from the all-time high, aligned to (any guesses?) 72 degrees. 7728.15 points of destruction over 73 long months yields a VR of 105.865/W, and dividing by tan72 (3.0777) gives the 45 degree scaling ratio (1×1) of 34.3976, which rounds up to 34.4/W. (Refer to the procedural explanation at the end) The initial vector circle provides resistance (AGAIN!?!) for current price. A 45 degree angle from the lower left corner provides general ascending resistance, as well as ‘squaring out’ with the vector high in-between the weeks of Jan 23 & 30. Price and time meet here again.

Interesting to note that the first pullback in the up trend on July 6, 2009 is 18 weeks from the vector low (in degrees: 18 + 72 = 90), and squaring with the vector low and careening 45 degrees NE gives proper support for the important Oct 3 2010 bottom. Once again, a 45 degree channel rises and gives effective S/R, offering a glimpse of possibilities if resistance does not hold.

In the next weekly Dow chart, the sine ratios of the entire pentagonal angle series (18, 36, 54 & 72) are used for S/R levels in the vector price range. The ratios are as follows: .309, .5878, .809 and .9511 respectively. Retrospective confirmation of price’s involvement with these levels is given at points A & B, clearly defining the first rally in the initial down trend. And, as expected, they define three very important top areas, which are spaced according to the sine functions of pentagonal angles!?!

In addition, the two major tops are at the intersection areas of the 45 degree angle from the vector’s SW corner. Both the major bottoms occur either at or above the intersections with the lower 45 degree channel, with the next points being Aug 27 – Sep 3, 2012, and Apr 8, 2013.

And finally, we narrow down to the daily chart. The May 2 high of 12,876 to the Oct 4 low of 10,404.50 is 2471.50 points over 108 trading days (yes, 108!!). We get an initial VR of 2471.50 / 108 TD = 22.8843/TD, and then divided by tan72 (3.0777) we arrive at 7.4355/TD for 45 degrees and our scaling ratio. (Again, check out the procedure section at the end.)

Our last Dow chart clearly shows the sine of 72 at 12,755.04 retrospectively giving resistance in the initial down trend, as well as holding for current price (as of Thursday, Jan 26). A reversal candle pattern is forming at 45 degrees from the square point of tan36 (.7265) as well as from 72 degrees up from its inverse point (1 – .7265 = .2735) which picks off the high. Intersections for timing points work as well. This creates a massive focal point, seen through pentagonal geometry and trigonometry.

Conclusion: Monthly, weekly and daily pentagonal analysis provides a clear reversal area in price and time. As stated before, this is in no way a guarantee, just a high probability. If the Dow manages to take out all these harmonic relationships, there must be a LOT of force underway. In that event, circular and channel geometry can help guide the way.

And now for dessert: here’s a 32 year gold pentagon.  Interesting to note that the pentagonal (5-sided) width of 470 months x the square root of 5 (2.236) is 1050.95. When added to the vector high of 873, gives 1923.95, exactly 25 cents above the all-time high of 1923.70. Sound odd (pun intended)? 25 is 5 squared! Goodnight.

Method for drawing the pentagon: The vertical scale is adjusted so it is seen at 72 degrees (this is easy to do after the 45 degree vector ratio is calculated, as shown below, by using a right-angle tool on the screen, if your charting program does not have a locking ratio feature. The calculations for the vector ratios necessary are shown below.

(Procedural note: If performed in a traditional geometric environment (an ideal scaling ratio of $1 week, or 20 points/month etc.) a pentagon drawn from this vector would not be set at 72 degrees from the vertical, would not point straight up, and therefore offer different points in price and time, however certainly effective. In other words, this vertical pentagon is an
actual perfect geometric expression of this very down trend. There-in lays its power. I refer to this type of approach as ‘Relative Charting’, since it is price action itself that determines the actual price to time relationship (scaling ratio), as seen from 45 degrees.)

The initial down trend’s vector ratio (‘VR’) is:

72 degree VR = 4552.79/33M = 137.963/M

(Interesting to note that this figure is almost exactly 138, which is 72 + (33 x2), as in 72 degrees and the vector’s 33 months, as if price is telling us that 72 and 33 are very important)

45 degree VR = 72 degree VR / tan72 = 137.9633 / 3.0777 = 44.827/M

(Also worth noting is that the 45 degree VR is almost exactly 45!?! so far so good)

Next, the base length can be determined by calculating the vector length as a hypotenuse of a right triangle, which involves converting the price movement of 4552.79 into months by dividing it by the scaling ratio of 44.827, resulting in 101.56M (side a), and combining it with 33M (side b) using the Pythagorean Theorem:

(101.56)(101.56) + (33)(33) = 11404.16 = (106.79)(106.79)

The length of the hypotenuse is 106.79M, rounded up to 107 (almost 108).  So draw a horizontal line 107 months from the vector low to Sep 2011.

From there, draw upward 33 months to the right, and connect with the vector high of 11,750.28. This naturally gives you a 72 degree line. The pentagon is technically 172.79M across (33 + 106.79 + 33), and rounded up to 173M, at June 2014.

Next, calculate the pentagon’s top price by either dividing the vector range by .618 (giving 7366.974), or multiplying it by the Fibonacci expansion 1.618 (giving 7366.41). They are slightly different results, yet I see the logic in both approaches. The monthly charts presented opt for the first approach, which is ‘top-down’ (so to speak) and yields a price of 14,564.50.

Connect both the vector high and its right-sided counter-part point to this top price at ½ of 106.79M from the vector low which is 53.40M (rounded), right in-between March and April 2007.

For the external circle, locate the center of the pentagon by multiplying the height by the center ratio of .4472, and add it to the vector low:

7366.974 x .4472 = 3294.51 + 7197.49 = 10,492 center height

Draw a circle from this center until it conforms to the pentagon’s points. Also, try an internal circle, which touches the mid-points of the sides.

How would you describe your job?

I advise clients regarding their financial investments based on my research and beliefs about where the markets will go in the future. 

What led you to look at the particular markets you specialize in instead of another tradable?

I very much enjoy following all markets, as they move in different directions. However, I especially like precious metals which represent “real” assets and currencies whose fluctuations reflect the always-changing worldwide economic conditions.

Do you look at any fundamental or economic inputs to develop your opinions?

I mostly rely on technical patterns and signals which I believe provide a more disciplined guidance of market entry and exit points than fundamental analysis and valuation.

Can you share any longer term market opinions?

Since the 2007-2008 financial crisis, the deflation scenario is well established. From time to time, big injections of money by governments cause a countertrend to this scenario. In deflation, all assets are liquidated to pay debt. The safety of short-term treasury bills for a large portion of assets is mandatory, and the leftover portion could be used for capitalizing on the downtrend of stocks and commodities. For the last 10 years, the stock market remained largely unchanged, but gold and silver (representing “real” money) greatly appreciated. Therefore, gold and silver should make up 5% of a portfolio.

What advice would you have for someone starting in the business today?

I believe education is crucial, an MBA and/or a Masters in Finance, as well as the CMT Designation which I think was extremely helpful in helping me learn about technical analysis in a structured and comprehensive manner. I definitely recommend everyone who wants to trade the financial markets professionally to take part in the CMT Program. In addition, it is imperative to follow the markets that interest you regularly, and take the time to analyze the charts’ various patterns over both the short and long-term. The most important is to develop your skills in such a way that it builds your confidence in approaching the markets.

Most trading strategies have a common goal, to reach the successful conclusion of a bet and walk away with a profit. But while the position is open it is at risk and stays that way until closed anywhere between full profit and full loss. Accessing other people’s risk, where they got in and how much of it is underwater, is the ultimate leading indicator. Unfortunately, besides the risk and back-office team of big brokerages and exchanges, few have access to this gold mine of data, and none can legally use it to trade (Peck, 100).

But how about inferring that information by overlaying basic trading strategies over market movements? This isn’t about unearthing profitable strategies, at least not directly, but to reverse engineer the market’s past intentions to induce future ones.

The idea of watching popular market strategies to anticipate crowd behavior isn’t new, the 200 day moving-average, amongst many others, springs to mind:

“First I will look at the 200-period EMA… This is the strongest moving average and represents serious support or resistance – it was enough to stop the crash of 2000…” (Carter,334).

The strategy for this indicator needs to be simple and used, in full or in part, by as large a pool of traders as possible. It also needs clear targets to delineate profitable from unprofitable data.

An approach with a one-to-one risk-reward ratio, by entering long on every bar that moves above the prior one and short when moving below, is a good starting point. This ‘1:1 RR’ strategy is in essence a basic continuation trade. Whenever a new bar goes above the previous one, a long position is initiated a tick above the previous bar’s high. The same goes for shorts, initiated at a tick below the previous bar’s low. The profit target is the length of the signal bar from the entry price, and the risk is that same  length but in the opposite direction from the entry point:

The ‘1:1 RR’ strategy can be applied to any product and chart type. But the time frame should be geared towards popular ones such as the 3, 5, 15, and 60 minute (Brooks, 6), to ensure that as many traders are watching the same setups. All the following examples use the well traveled, five minute chart of the SPDR S&P 500 ETF. It is there that I have often seen ‘1:1 RR’ trades fail to reach their profit targets by a single tick. This is the confirmation that enough traders are paying attention to this strategy to make it a worthwhile indicator candidate.

Long Trade:
Entry = CURRENT PRICE > HIGH (previous bar)
Profit = CURRENT PRICE > HIGH (signal bar) + (length of signal bar)
Stop-Loss = CURRENT PRICE < LOW (signal bar)

Short Trade:
Entry = CURRENT PRICE < LOW (previous bar)
Profit = CURRENT PRICE < LOW (signal bar) – (length of signal bar)
Stop-Loss = CURRENT PRICE > HIGH (signal bar)

The indicator holds all active trades, collects new trades as they trigger and drops those reaching their profit or stop-loss targets. From that data it tallies and displays the current level of profitability of the bulls and the bears.

Chart 1 – SPDR ETF – 12/01/2011

Chart 1 shows the strategy in action with the blue line representing the bulls and the red, the bears. The profitable trades are above the zero line and the unprofitable ones, under. The indicator shows that trades triggered using this ‘1:1 RR’ strategy spend more time in the unprofitable territory under the zero line. Yet this isn’t important as we aren’t trading the strategy, instead we want to determine which side controls the market. From a technical analysis perspective, 1 PM shows the bears’ failed attempt at breaking below the day’s prior low. The market forms a double bottom and reverses upward. The indicator confirms that the bulls are dominant during that period as the blue line is above the red one and, at times, above the zero line.

As the title suggests, the focus of this article is on risk. While a ‘1:1 RR’ trade is active (hasn’t hit its profit or stop loss target), it is made entirely of risk. But risk can be broken down into two types, a profitable portion equal to or better than break even and an unprofitable portion worse than break even. To render this indicator more sensitive, we update the strategy to only use unprofitable risk. The assumption is that loss will be dealt with aggressively and lead to action. We therefore add another condition to the profit formula to ignore trades that reach positive risk. So when the low of a new candle bar is higher than the entry price in a long trade, we drop it from the indicator. Similarly for short trades, when the high of a candle bar is lower than the entry price, we drop it. This is like having a stop on the trade to close the position once it hits or exceeds break-even.

Long Profit:
CURRENT PRICE > HIGH (signal bar) + (length of signal bar)
or
ENTRY PRICE < LOW (current)

Short Profit:
CURRENT PRICE < LOW (signal bar) – (length of signal bar)
or
ENTRY PRICE > HIGH (current)

As the updated formula now only returns negative values, we can plot the data differently to makes things easier to read. First we use bars instead of lines and, second, we flip the scale on the bulls. The scale of the bullish bars starts at zero and increases with risk while the bearish one starts at zero and decreases with risk. The bulls are therefore restricted to the top half of the indicator, the bears to the bottom half, and both sides start at zero.

Chart 2 – SPDR ETF – 12/20/2011 1

Chart 2 shows the bulls piling on risk (blue bars going higher) and the market rewarding them by climbing higher. The red bars are short and farther apart, corroborating a market controlled by bulls.

Chart 3 – SPDR ETF – 09/26/2011

Chart 3 shows the first pattern gleaned from the indicator. The day opens with the bears in control, then the bulls, then the bears and finally the bulls again. This behavior is clearly reflected on the indicator as it parallels the market closely. When the indicator shows a clear and gradual accumulation of risk on one side and thinning on the other, the market usually follows the risk.

Chart 4 – SPDR ETF – 12/13/2011

On Chart 4, a tug of war happens between bulls and bears between 1 PM and 2 PM. Both sides hold risk so the market could go either way. Around 2:30 PM, the fate of the market is sealed as it drops and the bullish risk evaporates.

One last tweak to the ‘1:1 RR’ strategy will clarify the indicator even further. We now exit trades as soon as a counter movement contradicts the entry bar by at least one tick. So if the trade is currently long and the following bar goes below the previous bar by one tick, the trade is stopped out. Same goes for shorts, a bar that goes 1 tick above the previous one closes the short trade. This not only emulates an automated trailing stop-loss, it adds an additional weakness detection to the indicator.

Long Stop-Loss:

CURRENT PRICE < LOW (signal bar)
or
CURRENT PRICE < LOW (previous bar)

Short Stop-Loss:

CURRENT PRICE > HIGH (signal bar)
or
CURRENT PRICE > HIGH (previous bar)

Obviously, patterns don’t always work. A failing first pattern (i.e. risk failing to move the market accordingly) is the second pattern. If risk points towards a direction and the market ignores it, we have to anticipate that those unfortunate trapped traders will eventually dump their holdings. If a lot of them are caught on the wrong side of the market, the reversal can be violent.

Chart 5 – SPDR ETF – 12/09/2011

Chart 5 shows the second pattern in action. Right before 11 AM the indicator shows a strong bearish spike at the end of the bearish leg. Spikes can hint at exhaustion and trapped traders. On the above chart, the spike marks the exact delimitation between the bear and bull leg.

This indicator isn’t meant to be a stand-alone tool as the signals aren’t smooth and bounce from one extreme to the other fairly rapidly. By combining this indicator with other trading methodologies, it can offer additional insight during crucial trading decisions. The following section will look at how this indicator and its two patterns compliment the usage of basic trend lines and trend line breaks.

Chart 6 – SPDR ETF – 05/17/2011 to 05/19/2011

Chart 6 shows the market attempting to break down late in the morning on the 19th but failing to pierce the long-term bullish trend line. The indicator confirms the bulls’ overwhelming presence around that time.

Chart 7 – SPDR ETF – 06/02/2011

Chart 7 shows two failed attempts by the bears to break the bullish trend line (arrows on the chart). The second failed bear spike can be interpreted according to the second pattern where an exhaustive bearish spike ends the leg.

Chart 8 – SPDR ETF – 05/252011 to 05/26/2011

On Chart 8, a bearish spike on the indicator around 10:30 AM is a warning that the bear leg (and multi-day trend line) may soon end.

Chart 9 – SPDR ETF – 06/29/2011

Chart 9 shows a short bearish leg on the open and a bear exhaustion spike on the indicator. The bulls come in around the trend line and succeed at moving the market upwards.

Chart 10 – SPDR ETF – 11/08/2011

Chart 10 shows bears giving up right before midday with an exhaustive bear spike. The bulls subsequently move in and break the trend line.

Chart 11- SPDR ETF – 05/04/2011

The bearish trend line on Chart 11 is untested until midday. At 1 PM, the bears fail to go lower and form a small double bottom. The indicator shows a strong effort by the bears yet the market reverses and dries up all bearish risk for the following hour.

Chart 12- SPDR ETF – 08/23/2011

Bulls take over the open of the 23rd on Chart 12 and break a bearish multi-day trend line. The indicator shows both patterns, a strong bearish spike at the open followed by a stretch of bullish strength. A similar situation is seen right after 2 PM which carries the market upwards into the close.

Final words:

Designing an indicator from an existing strategy can yield additional information that may not be as obvious from candle data alone. This custom indicator seems adept at highlighting both continuation patterns, usually in an overwhelming one-sided imbalance, and reversal patterns, in the form of failing risk and exhaustive spikes.

Lastly, the ‘1:1 RR’ continuation strategy is one perspective, others, such as counter trend and mean reversal, may prove just as interesting.

References:

• Peck, Sarah. Investment Ethics. Wiley, 2010.
• Carter, John. Mastering the Trade Proven Techniques for Profiting from Intraday and Swing Trading Setups. McGraw-Hill, 2005.
• Brooks, Al. Reading Price Charts Bar by Bar: The Technical Analysis of Price Action for the Serious Trader. Wiley Trading, 2009.

Sherpa Funds in Singapore is a small team with over 50 years trading experience, spread between London, New York, Tokyo and Singapore. We have worked for Credit Suisse, Barclays Capital, Bankers Trust and TD Securities and dealt in Foreign Exchange, Interest Rates, Emerging Markets, and Credit products. In 2010 we decided to channel our experience to create proprietary trading algorithms for liquid markets.

Primarily we wanted to eradicate human emotion from the whole trading process, but of course we also wanted to maximize long term growth, minimize negative return volatility and reduce the occurrence of large downside losses. We focused our back testing from 2003 on “liquid markets” such as G11 currencies, equity indices, interest rate futures and a handful of globally liquid commodities. We programmed all kinds of traditional forms of technical analysis (including DeMark indicators), mathematics, historical volatility measures etc. and began writing and back-testing rules. We quickly found some very interesting results and observations.

Firstly of the four components to a trade; entry point, stop loss, take profit and position sizing, it was position sizing that we found to be BY FAR the most critical factor, not only for consistent returns but also in avoiding crippling losses. We already expected this having discussed the Kelly Criteria in detail so we went about position sizing with a series of factors most notably using historical volatility and expectations of returns.

We were also surprised to see how the market dynamics changed throughout the day. We have a very consistent short term rule (which lasts hours not days) but we found that the results behaved very differently depending on what time of day the trade was entered. Back testing showed the trade made money for 15 of the 16 hours between the London open until the New York close. But then for the subsequent 8 hours in the Asia-Pacific time zone the rule was random at best.

This was consistent for the 8 years of back testing, and we put this down to the fact that London and New York markets are much more liquid so move in traditional waves and patterns, whilst in Asia-Pac (and especially between New York close and Tokyo open) the illiquid market is driven much more by large one-off institutional flows and plain noise. The result table is shown below based on the Singapore time zone, New York closes at 0500 in the morning and London opens at 1400.

We also found some interesting market dynamics within the “Risk-On” and “Risk-Off” scenarios. We all know that equities, commodities and FX “carry-trades” grind higher during the accumulation and participation phases but fall much more aggressively as the herd rush for the small exit, “going up the escalator and down the elevator”. So programs have to be tailored differently to rising risk markets as opposed to sell-offs. AND we also found as soon as the S&P trades over 4% below the 200 day moving average, traditional forms of technical analysis become less useful as the market takes on more powerful and exaggerated “panic like” characteristics.

We are also nervous of the correlation of markets during the waves of “risk on” and “risk off” so we have been careful to target a variety of products over a number of different time frames and we analyze the correlation of the outcomes within our back tested rules, and between the rules themselves. Below is a graph showing the back testing of 2 short term rules that are currently live in the Sherpa Funds. The testing shows the results are uncorrelated to the S&P index, are consistent and that negative volatility is very low. The portfolio ratio of negative day volatility to returns is half of the same ratio for individual rules.

Of course back testing has limitations and we have to be aware of possible data-mining. But since we went fully live on the automated trading we have only had positive months ; October +1.9%, November +1.5% and December +0.2% trading with our own money to create a track record to market at a later date. So far we appear to be heading in the right direction.

The MTA Educational Foundation (MTAEF) would like to welcome Marcelo
Kim, the first intern selected in the MTAEF’s new Internship Program. Over
the next four months, Marcelo will be shadowing various technicians in Boston and New York. If Marcelo contacts you, the MTAEF asks that you please try to accommodate him in your schedule! Below is an introduction from Marcelo:

“My name is Marcelo Kim and I am a recent graduate of Baruch College. At Baruch, I majored in finance and was a member of the Golden Key Honor Society. My past work experience includes: working as a business analyst intern at AXA Liabilities Managers and as an intern at a local accounting firm. I am an avid fan of all sports and also played tennis for the Baruch team. You would be surprised to know that I speak Portuguese fluently as I was born and raised in Sao Paulo, Brazil!

I was introduced to Technical Analysis through a class taught by Bruce Kamich, CMT. During the first lecture, Professor Kamich said he would use the F word occasionally. His statement was puzzling at first, but I later learned that it stood for fundamentals. Initially, I thought fundamentals were the only method used in investments – however the more I learned about Technical Analysis, the more appealing it became to me. I believe this opportunity of meeting market technicians will help me learn more about the implementation of the technical side in investments in a variety of markets.”

Michael W. Covel is the author of “Trend Commandments” and “The Little Book of Trading.” Mr. Covel’s  past bestsellers include “Trend Following” & “The Complete Turtle Trader.” His trend following books have been translated into over 10 languages.

In 2009, Mr. Covel directed the documentary film “Broke: The New American Dream.” He has presented live to audiences in Chicago, Dallas, Hong Kong, Las Vegas, Macau, Miami, Paris, São Paulo, Tokyo and Vienna. Mr. Covel’s research firm (www.trendfollowing.com) has been training trend following traders since 1996, across over 70 countries. His original website was TurtleTrader®, which has drawn millions of visitors.

Mr. Covel holds a B.A. in politics and government from George Mason University and an MBA from The Florida State University.

A note from Michael Covel:

“For 15 years, I have pushed hard to define trend following standards through books, film, web, and live presentations. Has it
worked? My first book Trend Following is required reading for all employees at numerous billion-dollar hedge funds, but many of those funds only admit their support off the record. Silent support? Check. Gatekeeper? Check. My original plan did not involve becoming a trend following linchpin, but I am very passionate about my role.”

Michael Covel will be presenting at the 2012 Annual Symposium on Thursday, April 19th, 2012 at 3:45 PM. He will be a keynote presenter, discussing “A Trend Following Wake-up Call: The Search for Return – The philosophy, psychology, and results…when you don’t try to predict.”

Michael previously gave a presentation on “Trend Following” in the MTA’s Educational Web Series in November of 2009. A copy of this presentation can be seen online by visiting http://go.mta.org/223.

For more information on the 2012 Annual Symposium, or to register, visit http://symposium.mta.org