The information in boldface below was updated on July 13, 2008.

HUI, the Amex Index of Unhedged Gold Mining Shares, has completed a bearish euphoric topping pattern stretching for two years, as follows:

First left shoulder: 401.69, May 11, 2006.

Second left shoulder: 402.27, September 21, 2007.

Third left shoulder: 401.50, October 1, 2007.

Left weak near-peak: 463.06, November 7, 2007.

First blowoff top: 489.49, January 14, 2008.

Confirming left near-peak: 478.70, January 30, 2008.

Penultimate top: 505.65, March 4, 2008.

Final euphoric zenith: 519.68, March 17, 2008.

Matching right near-peak: 475.60, April 16, 2008.

First weak right near-peak: 459.48, May 21, 2008.

Second weak right near-peak: 466.26, July 1, 2008.

Third weak right near-peak: 455.57, July 11, 2008.

It is especially interesting to notice that GDX, the most popular fund of gold mining shares, showed nearly a perfect bearish topping pattern with its 53.84 peak of November 7, 2007 being matched by a slightly higher high of 54.23 on January 14, 2008, yet another peak of 55.20 on March 4, 2008, and finally a fourth phony upside breakout at 56.74 on March 17, 2008. While professionals and insiders were increasingly selling into each high, amateurs were tripping over themselves to be as fully invested as possible, so as not to "miss out" on supposedly unlimited additional gains that never materialized. Since GDX represents a real fund that people have been buying and selling, it is astonishing that a more than $200 rise in the gold price from early November 2007 through the middle of March 2008 was only able to push GDX less than 5.4% higher. This powerfully negative divergence is sending an ominously bearish message for precious metals over the next several months, which is confirmed by the early May multi-month lows for both HUI and GDX.

The biggest danger for gold mining shares is that there is no proven support level until these shares have fallen by more than 35%, and perhaps even more than 40%, from their March 17, 2008 highs. Although gold mining shares remain in a strong long-term bull market which began on November 26, 2000, the short-term danger is severe and has been overlooked by most participants.

One question that readers often ask is why many asset classes in recent years have set new highs or lows before reversing sharply in the opposite direction. The answer is simple and requires understanding the following two facts about the financial markets: 1) they will always act in the way which will benefit the fewest number of people; 2) if a significant number of traders recognize a certain established pattern, the financial markets will mutate like a clever virus in order for rule #1 to prevail. Therefore, since so many traders have become familiar in recent years with the concept and implications of a breakout, the financial markets will intentionally make numerous false breakouts to fool these traders, and will then make a sharp and devastating U-turn in the opposite direction. In recent months, we have seen exactly this pattern with assets as diverse as the S&P 500, gold mining shares, and the U.S. dollar, among many others.

Therefore, a double top in the modern era almost always sees a higher high than the previous peak, while a double bottom is almost always accompanied by a lower low than the previous bottom.

HUI is also completing a long-term bullish bottoming pattern with several nearly exactly repeating lows, as follows:

First low: 278.47, March 10, 2006.

Second low: 270.54, June 13, 2006.

Third low: 274.72, October 4, 2006.

Fourth low: 284.85, August 16, 2007.

Fifth low: ???.??, August/September/October/November ??, 2008 (295 estimated).

HOW LOW WILL HUI GO?: HUI is the Amex Index of Unhedged Gold Mining Shares. How low will HUI eventually go, exactly, over the next several months? My guess is 295, for the following reason:

One very useful guide is to observe that on December 2, 2003, HUI reached a peak of 258.60 which was not exceeded for about two years. As a rule, during its bull market which began on November 25-26, 2000, HUI has bottomed moderately higher than its second-most-recent high-water mark during each subsequent extended correction. Another guide is found by measuring the entire gain in HUI from its May 16, 2005 bottom of 165.71 to its March 17, 2008 peak of 519.68. If HUI surrenders exactly 61.8% of this increase--known as the key Fibonacci retracement--it would put HUI at 300.93. Usually, any Fibonacci retracement support level will be broken by a few percent to the downside. It is significant that the August 16, 2007 low for HUI was at 284.85, since this indicates that additional downside attempts through the summer of 2008 will likely head directly for this target. For more details on the new weighting for HUI, go to the American Stock Exchange's official web site at http://amex.com/ .

HUI CORRECTIONS IN THE PAST 4 YEARS: HUI, the Amex Goldbugs Index of Unhedged Gold Mining Shares, has experienced five corrections averaging more than 1/3 in magnitude in the past six years. I strongly believe that the sixth such correction began on January 14, 2008. Here are the precise dates and figures of each such pullback:

2002: 154.99 on June 4, 2002 to 92.82 on July 26, 2002: 40.1%.

2003: 154.92 on January 6, 2003 to 112.61 on March 13, 2003: 27.3%.

2003-2004: 258.60 on December 2, 2003 to 163.81 on May 10, 2004: 36.7%.

2004-2005: 248.18 on November 17, 2004 to 165.71 on May 16, 2005: 33.2%.

2006: 401.69 on May 11, 2006 to 270.54 on June 13, 2006: 32.65%.

2008: 519.68 on March 17, 2008 to ???.?? on ??? ??, 2008: ??.?% [I'm guessing the eventual low will be near 295, a decline of 43.2%].

It is important to note that the recent pullback for HUI from 519.68 on March 17, 2008 to 384.53 on May 1, 2008 was almost exactly 26%. The previous five corrections (see previous paragraph) had not experienced even a single decline which was so shallow, and therefore this strongly suggests that the 2008 downtrend for gold mining shares is far from over.

CONSIDER FUNDS FIRST: Many readers will probably be interested in purchasing funds of gold mining shares for some portion of their investment in precious metals, not wishing to assume the increased risk and volatility of owning shares of individual companies. There are several dozen such funds.

Since May 22, 2006, there has been a new exchange-traded fund of gold mining shares called GDX, which is intended to track the Amex-listed index GDM. More information on GDX can be found at http://www.vaneck.com/gdx. The management fee for GDX is currently only 0.55%. Since GDX has the lowest fee of all true gold funds, and can be traded intraday, it is currently my favorite gold fund. For Canadian investors, XGD on the Toronto Stock Exchange is a relatively close approximation for GDX, and also has a management fee of 0.55% (now that's collusion with a capital C!). More information on XGD can be found at http://www.ishares.ca/product_info/fund_holdings.do?ticker=XGD . However, XGD, while as equally diversified as GDX, generally trades with a greater bid-ask spread than GDX, thus making it comparatively somewhat more illiquid. Since the two funds will have an almost identical total performance when measured in any single currency, you should trade the one for which it is easiest for your broker to get the best fills.

Meanwhile, ASA is a closed-end fund of precious metals shares, including especially gold and platinum, which is more heavily weighted than most gold funds in South Africa. South African mining shares are currently heavily out of favor, which likely means that they will outperform other mining shares until more normal historic ratios are restored. Since the annualized management fee for ASA is currently 0.53%, which is slightly lower than the 0.55% fee for GDX or XGD, its similar ability to trade intraday makes it worth buying, especially whenever ASA's discount to net asset value exceeds 12%, which happens periodically. ASA is rather illiquid, so always trade it using limit orders rather than market orders. Check http://www.etfconnect.com/ or http://www.closed-endfunds.com/ for the latest information on net asset value, expenses, and portfolio holdings for ASA and other exchange-traded and closed-end funds.

All but three of the open-end gold funds charge more than 1.2% percent of the total assets each year as a management fee, in some cases more than two percent annually. Some of these funds even charge upfront or redemption fees. If you feel that a particular fund manager has a track record which justifies such a high expense ratio, then you may wish to continue to invest in such a fund. However, it is possible that such a manager may not continue his winning streak, or his success may encourage him to leave for another company or to start his own hedge fund. The subsequent management will not necessarily be as competent, and may charge a significant fee for switching out of the fund. Caveat emptor.

The two open-end gold funds which charge the most reasonable fees are BGEIX, the American Century Global Gold Fund (current annual expense ratio 0.67%), and VGPMX, the Vanguard Precious Metals and Mining Fund (current ratio 0.28%, although it is currently closed to new investment). I have a strong preference for BGEIX over VGPMX, for the following reasons: 1) BGEIX charges a redemption fee of 1% only if the shares are held for less than 60 calendar days. VGPMX charges a 1% fee if the shares are held for less than a year. 2) BGEIX contains all pure gold mining companies. VGPMX contains several energy and base metal producers. If there is a fear of a recession, or similar economic developments, energy and base metal producers will likely underperform ordinary gold mining shares. Besides, I do not want diversification if I am purchasing a gold fund; I want gold mining shares, period. 3) BGEIX has always been open for new investment. VGPMX sometimes is closed for new investment, even if one has a considerable current holding in the fund. Note: VGPMX is currently closed to new investment, even if you currently own shares of the fund. I love Vanguard's funds in general, including their exchange-traded funds, but two big thumbs down to Vanguard for that! So stick with BGEIX if you prefer an open-end gold fund.

Appendix: I currently do not recommend FSAGX, the Fidelity Select Gold Fund, as it no longer permits hourly trading, but only end-of-day trading; thus, its higher expense ratio of 0.97% is no longer justified. As readers who have been tracking gold mining shares for many years already know, being able to buy and sell at 10 a.m. is worth a lot, since it is around this time of day that gold mining shares usually make their lows when they are forming an important bottom. Sometimes HUI will rise several percent between 10 a.m. and 4 p.m., as it did on May 10, 2004 (when HUI made its nadir of 163.81). Once it is appropriate to sell gold mining shares, it is also usually advantageous to do so at 10 a.m., when they generally peak during the formation of market tops.

GOLD AND REALITY: Gold and gold mining shares often correlate closely to the real rate of return for short-term U.S. time deposits. The lower the real rate of return, the better for gold and its shares, and vice versa. The real rate of return is equal to the nominal time deposit rate (roughly equal to the anticipated Federal funds rate) minus the inflation rate. (July 13, 2008) At the current time, core inflation is about 2.75% while the anticipated Federal funds rate is approximately 2.20%, yielding a real rate of return of negative 0.55%. This rate has become less negative in the past few months after having previously fallen sharply since the summer of 2007. A special note to conspiracy fans who use an "underground" or "more realistic" rate of inflation: notice that this supposedly "better" rate follows exactly the same chart pattern as the official government rate. Therefore, the current spread had recently touched its highest point in more than six years, even if its absolute value differed from the numbers stated above. It should be noted that when gold experienced its lowest inflation-adjusted levels in 1999-2001, the inflation rate averaged 2.0% while the anticipated Federal funds rate was as high as 6.5%, yielding a real rate of return of as much as positive 4.5%. This positive 4.5% rate of return is the primary reason why gold fell all the way to $252 per troy ounce, not because of some ridiculous manipulation theory.

LOOKING FORWARD TO 2060: I'll make a very far forward prediction by stating that I believe U.S. equities will make a double bottom in 2010 and 2018, with the Nasdaq bottoming in 2010, more than doubling sometime thereafter, and then making a final double-bottom retreat in 2018, followed by a roaring bull market that will create a very long-term double top for the Nasdaq just above 5000 by perhaps 2037, followed by the next secular bear market. The all-time Nasdaq high of 5132.50 from March 2000 will probably not be seen again until around 2060, six decades after it had previously visited that mark, and obviously representing a much lower real level because of inflation. Jeremy Grantham, in the February 6, 2006 edition of Barron's, made the exact same timewise prediction: U.S. equities will achieve their deepest nadir in 2010. His logic is that any asset class typically takes about one decade to go from top to bottom (U.S. equities peaked in March 2000). U.S. residential real estate lovers, take heed: the peak in August 2005 [National Association of Realtors] implies that the bottom for U.S. housing prices will occur near 2015.

LONG-TERM GOLD OUTLOOK: Gold will continue to make a pattern of higher lows as its strong bull market from a nadir of $254.00 per troy ounce on April 2, 2001 to the present continues throughout the next decade or so. Important higher lows included $319.10 on April 7, 2003; $371.25 in the morning of Monday, May 10, 2004; $410.75 on February 8, 2005; $428.00 on August 30, 2005; $534.00 on March 10, 2006; $542.50 on June 14, 2006; $563.25 on October 4, 2006; $602.00 on January 6, 2007; $641.50 on June 27, 2007; and $651.75 on August 16, 2007. My guess is that the next important low for gold will be in the summer or early autumn of 2008, and that it will be within 5% of $720 per troy ounce.

YOUR TYPICAL GARDEN-VARIETY SEVERE BEAR MARKET BOTTOM: U.S. equities in general will continue to decline until the dividend yield on the S&P 500, (July 13, 2008) currently at a historically low 2.36%, is between 6.5% and 10.5%. Great bull excesses are usually followed by equally severe recessions. Additional information on U.S. dividend yields can be found at http://www.indexarb.com/dividendYieldSortedsp.html/ .