I started at William O'Neil + Company Inc in January 1996 and I took a hiatus beginning in January 2003 to travel the world. Therefore it made sense to have KPMG verify those seven years beginning with my first month of employment at William O'Neil+Co Inc. The verification was done in early 2004. Incidentally, 1996 was my second year of triple digit gains over 100%, with 1995 being the first. I then proceeded to make triple digit gains every year through the year 2000, thus achieving triple digit gains for six years in a row.
I always wanted to work for William O'Neil after learning that he and his protégé David Ryan had exceptional long term track records that outperformed the markets over multiple market cycles. I read William O'Neil's book How To Make Money In Stocks back in 1989 and it revolutionized my way of thinking. Consequently, 1991 was my first successful year investing in stocks. O'Neil's hybrid method of applying fundamental and technical analysis to find stocks with the best chances of outperforming the markets made sense.
So began my long-term study of what characteristics the top performing stocks have in common. I worked with William O'Neil on some of these studies during the six years I was with his firm. After O'Neil appointed me head of research, I worked on the 1998 Model Book project which examined top performing stocks from 1992-1998. To date, I have gone through nearly 20 market cycles, all the way back to the 1920s, to study the stocks which made huge gains. I studied each stock in detail to determine which fundamental and technical variables predicted success. And I came up with a set of variables that the winning stocks shared.
I also created, and then refined, a timing model so I would be on the right side of the market whether we were in an uptrend or a downtrend. It has never missed a bull or bear market and I have used it real-time, under fire since 1991, my first successful year in the market. In backtests, it has well outperformed the major averages since 1974, the first year of thorough backtesting. To ensure the robust nature of the model, I have also spot-tested the model during the period of the 1920s and 1930s, which produced impressive results. The systematic portion of the model is a statistical formalization of price/volume action within the major indices. The discretionary portion of the model observes other factors such as the behaviour of leading stocks and sentiment/psychological indicators.
I remember in 1999, a salesperson who was retiring and had been with William O'Neil + Company Inc since the late 1960s came to my office and gave me a large stack of original O'Neil + Company market calls that were dated from 1968 to 1999. I went through each one and charted William O'Neil's buy and sell signals on the market. I saw that O'Neil never missed a bull or a bear market. This confirmed the validity of O'Neil's method in using price/volume of the major indices to time markets, and further underscored the importance of my statistical formalization of the major indices.
All of these in-depth studies are largely responsible for the success I have had in the stock market since 1991. These are neither my rules nor Wall Street's rules, but rules that are based on how the market actually works and how top stocks actually behave. These findings are demonstrated in the bi-weekly reports found at www.gilmoreport.com.
1996
I made only minor headway with my personal account (PA) during the first quarter of '96 as the market was rather sideways and mostly trendless. Over the years, I have found such trendless sideways markets to be the most challenging because it is easy to get nickeled and dimed. Just make sure you don't get quartered.
In mid-March '96, I noticed a few high quality stocks breaking out such as Iomega (IOM). Bases are akin to a rocket launch pad for the stock. When buying a stock breaking out of a base, you want to see a strong, constructive base so your stock carries the best chance of a successful breakout. I used to carry around a hardcopy of the William O'Neil + Company Inc Dailygraphs so I could learn the difference between constructive and defective bases. General market action was a key variable. For example, in a weak, downtrending market, the strongest stocks would often form the left hand side of what would eventually become a constructive cup-and-handle or double bottom base. The strongest stocks act like springs. Once the weight of the market comes off, they spring forward, breaking out of sound bases they formed during the market correction.
As more leading stocks broke out of sound bases, I began to buy. I was on full margin enjoying the rally from late March into June, sitting in my typical 12 to 18 positions. Incidentally, I noticed my trading style in terms of position sizing and number of positions and risk levels has not changed over the years. It is independent of account size. At any rate, my account peaked at +72% year-to-date by the time the market topped in early June '96. Over the next several days, each stock I owned then began to hit my sell alerts, so I sold. I eventually found myself in 100% cash by mid-June. I had no idea the market was going to have a mini-crash but I always stuck to my rules. While I buy based on both fundamentals and technicals, I sell purely on technicals. Price is the final judge.
The NASDAQ Composite proceeded to drop -19.6% peak-to-trough. Many of the high growth stocks I owned had far more severe corrections since they had also well outperformed the market up until early June. But I was protected, because I was in cash.
After the market finished its correction, my timing model gave a buy signal shortly after on August 1, 1996. I also noticed stronger leading names just starting to break out of sound bases as the stronger names are often the first to break out. So I began to buy in earnest once again.
By December '96, the year 2000 stocks TSRI, ZITL, ACLY had been discussed at length as to how they could avert potential disaster when the clocks ticked over to January 1, 2000. I noticed all three of these stocks which were the major players in this subsector breaking out of proper bases. So I bought. The buys put me over the 100% mark and I finished the year +121.57%.
All it takes is one or two good home runs in any given year to make up for all the small losses. And odds are that at least one or two home runs will be hit in a given year.
1997
During the first quarter of '97, the market was in a downtrend. By April, some portfolio managers with whom I was speaking were ready to throw in the towel for the year as frustration ran high. I had stayed mostly in cash during this period as my timing model had been on a sell signal, and there were almost no stocks worth buying. On April 22, my timing model gave a buy signal, the first buy signal since early January, and I noticed fundamentally strong stocks in sound bases starting to break out.
I continued to buy top stocks as the rallied continued, selling the weakest names in my portfolio to make room for the stronger names breaking out. I was effectively force feeding any available buying power into the strongest names. The market continued on its uptrend into October. I noticed most all of my stocks hitting my sell alerts by October 17, just a few days before the massive sell-off caused by the Asian currency crisis. I had no idea this crisis was going to occur nor that the market would sell off so hard. But as I had always done before, I sold when my stocks hit their sell alerts. Thus, I was safely in cash a few days before the markets sold off. My drawdown off the peak was just -6.5% compared to the -16.2% drawdown in the NASDAQ.
Incidentally, my success-to-failure rate in 1997 was one of the lowest ever, with a success rate of just over 20%. Yet I was able to achieve a triple digit return, just barely (102% according to my accounting, 98% according to KPMG) because the home runs made all the difference. So the % of winning trades is perhaps the least important variable in this style of investing. The % gain variable is far more important. In practice, my success rate is usually up around 50% during bull markets.
1998
While the first quarter of '98 was very profitable, the months from July through early October were some of the most challenging. The market peaked in mid-July. My stocks hit their sell alerts so I ended up back in cash in July. The market then staged a feeble rally in September. Very few high quality stocks were breaking out of sound bases. Thus there was little to buy during that month. I remember, however, many investors buying that month assuming the rally was continuing. Then in October, the markets sold off very hard, demoralizing investors. Many had losses by that point year-to-date. The big market bounce on October 8 led to my timing model issuing a buy signal on October 14.
I also noticed a few high quality stocks such as EBAY breaking out of sound bases in the ensuing days. EBAY was a most interesting IPO. It had one of the best business models and had first mover advantage in its space, much like YHOO and AMZN. It came public on September 24, then proceeded to lose more than half of its value due to the nasty bear market that caused the NASDAQ to return -33.1%. This is a good reason why fundamentals are only half the story. No matter how great a stock's fundamentals, a serious bear market will usually drag a stock down, no matter how great it is.
When the market bottomed and my timing model signalled a buy, it only took a few days for EBAY to hit its buy alert. On October 26, it gapped up out of what I call a U-pattern. These rare U-patterns can be seen in the strongest of stocks. The stock is so strong that it's not going to wait to form a handle, and the length of the base is often 4 weeks or less. I bought my first position in EBAY on the gap up, then bought a second position as the stock bounced off its 10-day moving average. I have found that the strongest stocks often constructively trade around the 10-day moving average.
Meanwhile, many investors had been so demoralized by the brutal bear market that began in July, so were sceptical as the market bounced in October. I remember some stayed short the market into November. The market kept rallying, forcing those who were reluctant to buy, to either cover their short positions or admit their error and start buying. However, they were late. A lesson they learned is that the best stocks are sometimes the first ones to break out as they may offer the best gains after the general market signals a buy. EBAY was a good example of this.
The fourth quarter of 1998 turned out to be one of my most profitable quarters on record. I had the advantage of my timing model and also in understanding technology. This enabled me to sense which stocks had first-mover advantages in their space. This made them true market leaders. I screened for stocks with top fundamentals, then shortened the list by investigating each stock in detail. Only a few made the cut. I then put buy alerts on the remaining stocks. This is another example of where my strategy is a hybrid between fundamental and technical and backed by historical precedent analysis. It is entirely misleading to call the strategy "momentum."
1999
Part of successful investing is knowing when a fundamental part of the market changes. In the late 1990s, it was the earnings metric. Some stocks that made huge gains had little to no earnings. While earnings are one of the most important variables I use to gauge the potential of a stock, I realized that sales growth was a useful metric for such stocks as well as market perception about a stock that determined its success during this time. Understanding the fundamental story behind the stock, together with understanding how Wall Street perceived the story behind the stock, gave me an edge.
From this, I learned that markets sometimes change in subtle and not-so-subtle ways. While certain key fundamental and technical variables continue to work cycle after cycle which forms the core of my strategy, other variables have a limited life. It is up to the investor to follow the markets closely so they can see when new variables can be used to enhance profits as well as when such variables lose their predictive value. Be wary of black box methodologies that claim to be profitable without having to be fine-tuned. They may work for one or two market cycles but had better be fine-tuned to keep up with changes in the markets.
In the first quarter of '99, most of the stocks that triggered buy alerts were technology stocks, as the Internet was touching so many aspects of technology. I always do what the stocks tell me to do, so I was fully invested in the uptrend that led to the April 13 reversal day in the CBOE Internet Index, an index which is a good measure of performance in the Internet space. I noticed many Internet stocks had staged or were staging reversals after making huge gains. Also, the day before, on April 12, some stocks announced they were going to attach '.com' to their name. Some more than doubled in price on the announcement. This extreme buying struck me as some sort of temporary climax top for the Internet group. On April 14, before the market opened, I noticed many of the stocks in my portfolio were going to open down. After the first few minutes of trade, there was a noticeable weight on Internet stocks. I remember giving my trader a 'shopping list' of stocks to sell just several minutes after the market opened. These stocks proceeded to tank and the CBOE Internet Index overall finished the day down 9.6%.
Timing is everything especially when it comes to handling high octane names. They must be handled with care, especially if you decide to concentrate your portfolio in one sector. Had I not acted quickly when I saw the warning signs, I would have given back a much larger portion of the profits I had made during the first quarter of '99.
The second and third quarters of '99 were the most treacherous I had experienced since 1991, my first successful year in the market. It also represented the largest drawdown for my timing model of -15.7% and the largest for my PA of nearly -50%. 2Q and 3Q were treacherous because there was a high level of volatility in a trendless market. It was not until October that the market resumed its uptrend in earnest. I'm glad to say that periods such as 2Q and 3Q '99 are rare.
It is important to stick with a winning strategy, in good times and bad. I had lived with my strategy since 1991. This gave me the confidence to stick with it even during the treacherous 2Q and 3Q of '99. I never lost sleep during this period nor have I ever lost sleep over the market. The key is that I always understand why I'm making or losing money. My timing model sheds much light on the character of the market. If it is struggling, such as it did during 2Q and 3Q '99, it helps me realize that we are in an unusual environment. In this case, the market was both volatile and relatively trendless, the Achilles heel of trend following. Fortunately, as demonstrated clearly in Michael Covel's book Trend Following which I highly recommend to any investor, markets tend to trend more often than not, thus guys such as Bill Dunn and John Henry can remain successfully in business with exemplary 25+ year long-term track records and still be thriving. That said, periods of steep drawdowns are part and parcel of trend following. The key is to stick with the strategy in both good times and bad. The profits made during the good times more than make up for the losses during difficult, trendless periods.
2000
Celera Genomics (CRA) had made a critical announcement about their mapping of the human genome in late '99, sending biotechnology stocks roaring ahead. The group continued to make huge price advances in early '00. As with the Internets, many of which had no earnings, market perception played an important role in the biotechs, many of which not only had no earnings but also had no sales. I applied what I had learned about market perception with the Internets and applied it to the biotechs. This was key to my making a triple digit return in 2000, a year when the market averages such as the NASDAQ Composite were down nearly 40%.
After the March 10 market top, I was quick to reduce my holdings off margin and into cash as each stock hit its sell alert in the ensuing days. I did some infrequent, light buying for the rest of the year, so was able to preserve most of my profits.
2001
In February, I pyramided a short position in the QQQQs. I remember as profits were building, I started calling this my 'Modena trade'. A Ferrari Modena back then was the hottest new model and cost about $250,000 with the mark-up. Thus, a profit of $500,000 on a trade would pay for the car, after taxes. I finished the trade at a profit of over $600,000, but I never bought the car. I learned that even though I had always wanted to own a Ferrari, I didn't buy it because it wasn't the act of possessing the car that was important. It was the idea that I could easily possess it. So buying it became unnecessary. It brought me greater pleasure to keep the capital on hand for investing in the markets. I found this way of thinking is more sound than trying to make the market pay for one's expenses or luxuries.
In March, I reshorted my QQQQ position. I pyramided the position as before, and profits were over $1,000,000. The trouble is that I went in too heavily, too late in the trade. Then on April 5, the market gapped up fiercely and rallied the rest of the day. I finally closed my position at the end of the day. I lost just over a cool million. My profits on the trade totalled close to $100,000. Gil Morales came into my office to shake my hand and tell me that's one hell of a ride and that he was sorry to hear about my loss. I then recalled the Victor Sperandeo interview in Jack Schwager's Market Wizards where he talks about going into a bar and telling the bartender, 'I just made $100,000 today. I really need a drink.' So the bartender asks, 'Why the down face? Shouldn't you be celebrating?' Sperandeo replied, 'The problem is, I was up $800,000 earlier today.'
2002 to present
This was another uneventful year where I stayed mostly in cash. I remember many funds going under. While 2001 was a massacre, 2002 was equally brutal. Few were left standing. Once the NASDAQ Composite was off more than 70%, the market seemed to be excessively oversold. I decided to take a small position in the QQQQs for a long term play. I reasoned that the market could go lower but historically, had always resumed a strong rally after being so oversold. This was true after the panic of 1907, after the Great Depression when the market lost almost 90% of its value, and after other serious market setbacks.
So 2002 was profitable by a hair due to this trade and due to staying mostly on the sidelines safely in cash. I would have been nicely profitable had I shorted indices on any sell signals issued by my timing model. This would be true in prior and in subsequent years. Thus, this bias I had toward staying in cash during bear markets was replaced in 2009 with the proper action of shorting major indices on sell signals. 2008 was another good year for my timing model.
The timing model's historical average is well over +30% from 1974-2006 by going 100% long the Nasdaq Composite on a buy signal (B), 100% short the Nasdaq Composite on a sell signal (S), and 100% cash on a neutral signal (N). The year 2007 was its worst with a return of -10.9% due to a high number of false price/volume signals. That was also the year when there were many cross currents including the peaking of commodities, the end of the housing bubble, the early stages of a breakdown of financials as seen in the XLF index, and the beginning of the recession. Fortunately, years such as 2007 are extremely rare.
On balance, my timing model continues to keep me on the right side of the market cycle after cycle. For historical interest, here is my timing model in action during the great crash of 1929, the crash of 1987, and their aftermaths:
The compressed, sideways markets observed from January 2004- August 2006 brought new trading challenges:
But whatever doesn't kill you, makes you stronger. In late 2005, I came up with a major refinement to my strategy which enabled me to make my initial buy in a stock's base just before it breaks out, what I call buying "in the pocket." Not only does this refinement work today but also worked beautifully in prior markets during the '70s, '80s, and '90s. We have discussed this in some detail in prior Gilmo Reports found at www.gilmoreport.com. This technique has improved my investment performance during the brief uptrends the market has had since 2004 such as from September to November '06 and from September to October '07, when the window of opportunity was open. As the saying goes, if I only knew in the 1990s what I know now, this would have further improved my returns. I think this nicely illustrates that I will always be a student of the market, always learning, and hopefully always evolving.