Monday, February 27, 2006

The next 10 days

Well, I haven't done great work on this blog over the last 10 days. Actually, no work. System HAG didn't want to make any trades last week or today, so the results remain the same as in teh last post. I know that I might be repeating myself, but besides giving updates on the performance of System HAG, I want to:

1) Go more in depth into the theoretical framework of this behavioral/emotional trading system.
2) Share past results and lessons from System HAB (the predecessor to System HAG)
3) Give my overall view of the state of various markets from a behavioral perspective. Especially with regards to real estate (top of a peak), commodities (middle of climb), and technology (the beginning of the ascent)
4) Start trading System HAG in a separate trading account
5) Start using System HAG to determine entry and exit points for a long-term investment account.

Thursday, February 16, 2006

The first 10 days

Today the market broadly rebounded and System HAG sold it's holdings of Harmony Gold for a gain of 2.96% and bought nothing. Overall, the program has now made an average of .90% on 12 trades. Since each trade was for 25% of the portfolio, the program returned 2.7% overall during its first 10 days.

Here's to the staying on that pace.

Once thing that I'm thinking about: Google rebounded 7% today, and the program wasn't holding it. Unsurprisingly, yesterday was the first day that I started to think that I didn't want to hold Google as a trader. The ultimate point of this program is to buy all the speculative securities that no one wants to own at that specific time. More about this in the future....

Wednesday, February 15, 2006

Back to Earth

I'm still waiting to find time to really get this blog going...

In the meantime, updates from the last two days of Hurwitz System Alpha Gamma, which I'll now begin referring to as System HAG.

Yesterday, the program "sold" Valero, which was bought last Friday, for a loss of 1.66% (adjusted for the dividend paid that day). The program then rebought it at the same closing price of 49.75.

Furthermore, of the four buys made on Monday, Marvell was sold yesterday for a gain of 1.45% and Pan American Silver was sold for a gain of 2.83%.

Google and Harmony Gold were held over until today, and both ended up realizing losses: Google of .96% and Harmony Gold of 2.94%.

The only buy made yestereday, Valero, was sold today for a gain of 1.65%.

All together, adding the results of these 6 trades to the 5 trades from last week, System HAG has now made 11 trades which have had an average return of .71% each.

Going into tomorrow, the program wants to own Harmony Gold...

Monday, February 13, 2006

Loading up

OK. I promise to touch on a new subject tomorrow, but today I'm still just giving updates....

The broad market had a very rough day today. The most speculative markets--metals and technology--got demolished. The only long Hurwitz System Alpha had, Valero Energy, was flat on the day (how rare!). We'll hold it until tomorrow.

As for the buys today--Google finally makes the cut. The program buys Google, Marvell Tech, Pan American Silver, and Harmony Gold at the closing prices... Will the winning streak continue?

Tomorrow and afterwards, I'll begin calculating a cumulative return. I'll assume that every buy is for 25% of the portfolio if not indicated otherwise. If there are more that 4 buys, all will be given equal weight, unless indicated otherwise. So tonight we're 125% invested because of the continued Valero Holdings....

Friday, February 10, 2006

Perfection

Apple rebounded dramatically today as anxiety dissipated. Sold for a gain of +3.63%. The average is now at 1.49% for the 5 trades thusfar. 1.5% is the average gain per trade that I'm shooting for. Not bad for one or two days of work. Today the program buys Valero Energy at its closing price of 50.65.

In the next few days, I hope to have posts detailing the history of Hurwitz System Alpha. I'm also going to explain why this blog is called The Bigger Boom. [--blog name changed to Psychological Arbitrage in April]

Thursday, February 09, 2006

Textbook results

So today showcased the power of Hurwitz System Alpha. The three stocks that we sold yesterday--Ebay, Red Hat, and Harmony--were -2.37%, -1.72%, and +0.87% today. Meanwhile, the one stock we didn't sell, BHP Billiton, was +1.29% to close at 36.86, or -.38% for the two days. Overall, our average buy has returned .96%. There have been 4.

Today, the program wants to buy Apple at its closing price of 64.95. Google is just about a buy at 358.77, but because I want to buy this stock so badly at this price, I'm going to not buy it. See the logic? If I wasn't scared to buy it at the close, other people probably aren't either. Every buy should be in a stock where there's significant worry. For instance, there have been reports today that Apple's new iMac is having severe video problems. Perfect.

I want this blog to be so much more than just updates on the trading system. Got to find time.



The first follow up

This entry will be dated on the 9th, but is being written as a summary of activity on the 8th:

I'm still figuring out how to optimally use Hurwitz System Alpha. The quantitative part of the trading program uses past price movements to tell me investors in a security are likely to be anxious. When I've run historical data experiments to test the program, I've depended solely on the quantitative portion of the program. But I think there needs to be a qualitative decision making framework superimposed on the quantitative selections. That is part of the point of this blog. To track the development of this framework.

Today, the program selected nothing to buy. This is quite common.

Yesterday, I detailed the decisions surrounding 7 equities. Here are the results:

Google: not a buy. Next day: +0.32%
Ebay--barely a buy. Closing price:39.93 Next day: +1.40%
Yahoo--no Next day: -0.06%
Red Hat--purchase. Closing price: 27.04 Next day: +1.18%
Pan American Silver--no Next day: -0.26%
Harmony Gold--purchase. Closing Price: 16.9 Next day: +1.66%
BHP Billiton--purchase. Closing Price: 37 Next day: 1.65%

To simplify trading decisions for this blog, we will assume that all buys and sells happen at the close. This is a reasonable simplification because the trading program takes advantage of athe tendency for traders to avoid purchasing securities before the close. Furthermore, we will assume that all sells happen at the close too. If the stock goes up a reasonable amount, say 1%, we'll sell it after a day. Otherwise, we'll hold it for one more day. After a second day, the security will be sold no matter what.

Of the four purchased securities, three were sold today. Only BHP carries over to tomorrow. The average return of the three completed trades is 1.41%. BHP will be sold tomorrow and it's return will be included at that time.

It is interesting to note that of the 4 stocks purchased, BHP is probably the stock with the least percentage of its float in the hands of short-term emotional traders.

Tuesday, February 07, 2006

Hurwitz System Alpha wants to trade.

Today was an interesting day in the markets. Though the general market was not extremely volatile, commodities and metals, which are in the midst of a long-term bull market, had a considerable sell-off. Also some of the most successful technology companies continued to decline significantly in value.

Though I'm in the process of setting up an actual account to track the results of Hurwitz System Alpha, I thought I'd give a sample of buy signals today.

Google--not a buy, but almost.
Ebay--barely a buy. Closing price:39.93
Yahoo--no
Red Hat--purchase. Closing price: 27.04
Pan American Silver--no
Harmony Gold--purchase. Closing Price: 16.9
BHP Billiton--purchase. Closing Price: 37

It all begins.....

OK. So here it is. The first post.

This blog is meant to be an extension of all my investment ideas. I'm here to talk about market psychology and investor psychology. I also want this to be an intimate diary. I will expose myself--strengths and weaknesses equally. The goal is complete transparency of me, my ideas, and my actions.

In the coming weeks, I will lay out the theoretical framework of a trading system I've developed, Hurwitz System Alpha. The goal of this program is to purchase postions that traders are currently scared to own and sell them as the fear subsides. This fear need not be all consuming. Traders might simply want to avoid entering a situation that might make them scared. The thesis is that collective anxiety is a powerful, but subtle, market force that is consistent enough to be exploited.

For example, if traders are thinking that an instrument is attractively priced, but that other people are going to continue to sell it, they might put off their purchase until the price stabilizes. No one wants to start a trade with a loss. Similarly, if short-term traders think that an instrument has become attractively priced at the end of a trading day, the traders might put off a purchase until the next day to avoid overnight risk. Once again, the traders are avoiding the potential for a significant loss upon opening a position even though the trader thinks the instrument is attractively priced.

The decision to give up a slight average monetary gain for peace of mind is not an irrational act for a human being. But it is inefficient, economically speaking, to make choices that will lower one's lifetime utility of wealth. In an efficient marketplace, someone else will come along who's figured out how to efficiently balance risks and will grab the money left on the table.

I'm focused on the situations where the market might not be efficient. Situations where arbitrage might not be complete. As a general rule, the more speculative the market, the more attractive to emotional momentum traders. These are the markets that are most likely to have a large collective effect from anxiety of individuals. Speculative booms, short squeezes, panics. History has shown markets to be far from efficient during these easily identifiable periods.

Hopefully, this blog will eventually turn into a website, an investment partnership, or a newsletter. For now, it will serve as an outlet to help develop these ideas. Over the next months expect the following:

1) Trading of an actual portfolio using Hurwitz System Alpha.
2) An exploration of the theoretical foundation of behavioral finance, and potential market applications besides Hurwitz System Alpha. A bottom-up approach.
3) An analysis of the psychological state of macromarkets--real estate, commodities, technology, etc.--and how to best invest in these different markets.
4) Discussion of the current state of investment analysis and trading. A search for potential systematic errors by hedge funds, investment banks, individual investors, and other market participants. A top-down approach.

The ultimate goal is the development of a cohesive framework to discover behavioral inefficiencies and exploit them. This work has the potential to both stabilize markets and be quite profitable.