Signal Type	Trade Date	Index	Return since issued
		Nasdaq 100
		Russell 2000
		S&P 500

Cumulative Returns since First TimingCube Live Signal () as of

Index	Long Only	Long Only with Margin	Long & Short	Long & Short with Margin	Buy & Hold
Nasdaq 100
Russell 2000
S&P 500

Market Update

Stocks moved lower on the week despite the Federal Reserve's decision not to raise rates for the first time in over two years. The announcement had been widely anticipated, but markets sold off after the accompanying Fed statement cited remaining inflation pressures and left the door open for more rate hikes later this year. With its decision to pause, the Fed acknowledged that the economy is slowing. This could mean bad news for stocks in the second half of the year as an economic slowdown coupled with inflation would result in diminished corporate profits. As a matter of fact, inflation fears resurfaced early in the week when oil prices hit $77 a barrel after BP had to shut down an Alaskan field that accounts for 8% of U.S. production. In such an environment, it is not surprising that investors remain cautious. This is illustrated by the sharp bearish reversal that stocks experienced Wednesday. After an initial surge following a better-than-expected report from Cisco Systems, markets reversed course to finish lower on the day on higher volume.

For the week, the Nasdaq 100 and S&P 500 respectively lost 1.14% and 0.99%. The Russell 2000 fared worse, as it shed 3.17%. The Nasdaq 100 and Russell 2000 still rest below both their 50-day and 200-day exponential moving averages (EMAs). For its part, the S&P 500 remains above its 50-day and 200-day EMAs. Our Cash signal remains in effect.

Trend Timing School

Evolution

As promised when we implemented the current Cash signal on July 14, 2006, the TimingCube brain trust has been hard at work on adapting the Model to market conditions which we had not experienced live or in backtesting before. Today, we are happy to disclose the results and accordingly, we will let the numbers speak for themselves and keep the words to a minimum.

Before we get to the numbers, we need to set the framework within which we always operate:

• Stay true to our Trend Timing principles of delivering a 100% mechanical system
• Focus on mid-term trends for infrequent trading
• Not engage in curve fitting or other tempering with historical model behavior
• Insure that enhancements improve the entire backtested period (the only exception is the 15% Cash signals which slightly decrease historical performance for added downside protection)

We always place honesty and transparency first, and accordingly, unlike many of our less scrupulous competitors, we will not alter or hide our live signal history. All the signals we issued live will remain unchanged in the results we publish. However, from now on, the backtesting that we are all interested in is the one for the Current Model and accordingly, in the next few days, we will switch the Results page to reflect the updated Model. For posterity and all of us who like to keep records, we provide a permanent copy of the Original Model backtest.

For complete details of the backtesting of the Revised Model, including all the signals and cumulative performance since 1989 see the "Nasdaq 100 Historical Performance with Revised Model". Our Model is still primarily driven by the Nasdaq Composite index but we used the Nasdaq 100 results to make comparisons with earlier published backtests easier.

Rather than analyze the backtested trade history, most readers will be satisfied with the resulting performance statistics as presented in the table below.

Table 1: Comparative backtest performance statistics since 1989

	Original Model	Revised Model	Buy & Hold
Annualized Performance
	43.95%	48.75%	12.78%
Cumulative gain
	61,067%	108,993%	724%
Number of trades	52	53	NA
Number of Cash Signals	1	7	NA
Average Number of Trades/Year	3	3	NA
Number of winning Trades	40	48	NA
Winning Trades Ratio (without Cash signals)	76.5%	89.1%	NA
Average Signal Duration	124 days	121 days	NA
Shortest Signal Duration	3 days	2 days	NA
Longest Signal Duration	531 days	531 days	NA
Best gain on a single trade
For a Buy signal (Long)	236.7%	236.7%	NA
For a Sell signal (Short)	36.7%	30.8%	NA
Average gain on a single trade
For a Buy signal (Long)	24.8%	30.9%	NA
For a Sell signal (Short)	8.4%	9.6%	NA
Worst loss on a single trade
For a Buy signal (Long)	-7.0%	-4.3%	NA
For a Sell signal (Short)	-4.7%	-2.0%	NA
Maximum equity drawdown
	-22.8%	-22.8%	-83.0%
- The backtest period is from January 3, 1989 to July 14, 2006 - For the Original Model, trades since June 18, 2001 are the live trades - All figures are for the Nasdaq 100 index - All TimingCube Model figures are for a Long and Short strategy

Also note that the backtest of the Original Model does not include the 15% Cash signals which were only introduced on March 12, 2004 (see "Drawdowns and risk management" for backtesting we published then).

The most notable changes in the comparison is the increased annualized performance which is directly attributed to a higher Winning Trades Ratio and Average Trade Gain. The changes between the Original Model and the Revised Model are very small in nature, but large in how they deal with more recent markets. Not surprisingly, the signals during the 1989 to 2001 period have not changed very much (since this is the period our Original system was developed on). Most of the enhancements have a more significant impact on the more recent period. For example, performance in recent years, e.g. since 2004, has drastically increased to over 20% annualized as opposed to a slight loss with the Original Model. This is on the Nasdaq 100 which has been the worst index over the last few years. Other indices such as the Russell 2000 and S&P 500 have substantially improved as well. In fact, the Russell 2000 is up an impressive 22.9% for 2006 year-to-date and 91.7% since 2004.

Another interesting statistic we have added is the "Maximum equity drawdown" which incidentally has not changed with the current Model. This represents the largest drop in value from a high to a low, regardless of signal boundaries. The largest such decline was -22.8% which happened between May and June of 2001, which compares to the -17.2% we have experienced over the last stretch of signals. So while double digit drawdowns are not pleasant (especially for highly leveraged investors) they are part and parcel of investing in the stock market and should always be kept in perspective by keeping in mind that first and foremost we are Trend Timers to avoid the large losses which the markets deliver from time to time. The backtest serves us well by highlighting the fact that the maximum equity drawdown for Buy & Hold over the same period is 83.0%, registered between March of 2000 and October 2002.

FAQ of the Week

Question: Does the Fed really set interest rates?

No. The Fed (short for Federal Reserve Board) does not really set interest rates. It cannot dictate long-term interest rates which are set mostly by market forces, but can at best temporarily influence the direction of the changes.

The mechanics are as follows. In the U.S. interest rate decisions are taken by the Federal Open Markets Committee (FOMC) as it targets a desired interest rate (the Federal Funds rate which currently remains at 5.25%). This is the guideline rate for the shortest overnight loans between the Fed and financial institutions. The way the Fed works to achieve the target rate is by engaging in the open market for bonds and other debt securities. Through agencies like the U.S. Treasury or private companies the Fed can issue or purchase back IOUs, effectively injecting or withdrawing liquidity from the system. The beauty is that the Fed also has the option of simply creating the dollars it wants to inject (what used to be known as "printing" in the old days).

So, for example, when the Fed wants to push rates down it can buy back large sums of bonds from banks and other institutions. Large sums in billions of dollars per day are routine. The banks then have extra cash reserves which, unlike the IOU they sold back to the Fed, are not earning interest and must be loaned out as rapidly as possible. In a competitive market the way to issue loans faster is by offering competitive rates, thus the downward pressure the Fed intended to create. The rub is that just because there are extra cash reserves available, does not necessarily mean borrowers want it at the expected rate.

Warm wishes and until next week.

The TimingCube Staff