Wednesday, November 04, 2009

What Will Tomorrow Bring: Program Trading for Everyone!

When the stock market (and every other market) is on a one-way trip to the clouds, brokers convince their prey (err, I mean customers) to "buy and hold" via "dollar-cost-averaging" with some nice "rebalancing" from time to time. As the great investor Issac Newton explained: What goes up must continue up. It's simple physics. It's the laws of Mother Nature. Don't you like nature? Of course. In that case, they might have a hot tip on a cool green stock you should sink your children's inheritance into.

Or you could just discipline yourself to sock away 10% of your monthly income in Spiders and forget about it.

Or you could get all control-freak and OCD and get all into options and day trading and lose your shirt.
These days, the markets follow a different law of physics called the Super Rubber Bouncy Ball law. What goes up hits the ceiling, accelerates, shoots downward, hits the bottom, accelerates further, shoots upward, bounces off the furniture and pings sideways into Mom's expensive vase. The vase shatters on the floor and scares the cat, which shreds the nearest curtain and then pisses on the carpet.

In such a world, the only way to grow your kids' inheritance is to be able to make money no matter what direction the market is moving. The only way to do that is to buy and cover low, sell and short high ... and play volatility in between ... all the while hedging against soaring interest rates and inflation.

In short - do what hedge funds and Wall Street have done for years ... Hopefully, with a little more prudence since it's your own dinero ... and maybe within the law.

I suspect you won't have armies of traders and analysts to do this for you like Wall Street does. Fear not. Hedge funds don't have armies either. You just have to learn to invest like they do: with computahs. Over 70% of all US stock trades are executed by computers following rules with no human intervention. Welcome to Program aka Algorithmic Trading. In short: rather than picking specific investments or trades, you define rules for investing and trading. Then the computers do the grunt work of hunting for opportunities that fit your rules and they execute the trades for you.

Yesterday: Have lunch with your broker. Listen and nod for an hour as he ruminates, boasts, rattles off arcane irrelevancies, pretends to see the future, and slips in stock pick recommendations. Then blindly agree, give him a check for another ten grand, and let him pay the lunch tab.
Tomorrow: Have a pre-defined amount of money automatically sent from each paycheck to your brokerage account. Have lunch with your broker. Listen and nod for an hour as he ruminates ... well, you know, they're creatures of habit ... but the difference is that his suggestions should be strategies (aka packages of rules that make sense together) as opposed to prognostication and specific picks.

Yesterday: Pore over stock info online. Read the WSJ. Check Morningstar and Jaywalk. Watch CNBC and Kramer (yuck!). Every day or two, jot down a stock symbol that someone seems to think are going to "pop." Then, log into eTrade or Merrill Online and key in your stock symbol, amount, and price. Then, sit back and watch as that stock inevitably does the unexpected. Finally, freak out and sell the stock. Take the money that's left and repeat.
Tomorrow: Pore, read, check and watch, but much less frequently. Every month or two, jot down a strategy that fits the new market and economic realities. Log on and key in the rules comprising the strategy. Have a pre-defined amount of money automatically sent from each paycheck to your brokerage account. Then, sit back and watch (or not) as the computer buys and sells things you didn't even know existed at prices you don't care about. Some trades go the right way and the computer exits with a nice profit for you. Other trades go the wrong way and the computer applies your stop-loss rules to ensure you don't end up living in a cardboard box. The computer repeats tens, hundreds of times a week without your input. Hopefully, profits pile up. If not ... well, bad strategy. Just like LTCM.

Don't look now, but the future is already here. Folks have been developing trading rules for decades. A growing number of folks are using a growing number of online brokerage sites which offer rule-based trading. It's a bit amateurish still, but getting better ... fast. This time last year, it was a treat to enter a rule requiring more than a stock symbol and a price. Now, the likes of TD-Strategy Desk and Credit Suisse-AES have rolled out platforms allowing folks to effectively build entire entry and exit strategies, taking into consideration multiple asset classes, company fundamentals, and market conditions.

Vague and arcane, right? Here's a simple concrete example:

Assumptions: Short-term, the dollar will continue to slide against Chinese and Emerging Market currencies. The US will buy less from abroad. Emerging market stocks will do better than domestic ones. The US stock markets will be volatile around earnings seasons. Medium-term, drug companies will be pressured by Obama to cut costs, and profits will shrink accordingly. Long-term, house prices will go back up while commercial real estate prices will go down.

A strategy might include a set of rules like:
  • Constantly assess cash available and asset allocation
  • Constantly assess current interest rates, dollar prices, relative yuan prices, etc
  • Constantly assess prices of bazillions of assets and fundamentals of bazillions of companies around the globe
  • Always maintain a certain cash balance, subject to a hard minimum plus some extra when market volatility increases ... and at Christmas times
  • Slowly build up interest bearing assets which mature in about June 2012, because that's when you'll be ready to buy your next fancy new car
  • Always maintain asset allocations within a few percentage points of the target pie-chart you've defined. If this appears to be limiting your profits because of missed opportunities (meaning another rule would have fired in the past and would have made you money, but it was prevented from doing so because of the allocation limits), alert you to reassess the allocation strategy
  • Auto-buy into dips in the market, especially around earnings seasons
  • Auto-sell when the asset reaches a relative target, either a % gain or crossing a certain statistical threshold like a moving average
  • Auto-exit if any trade goes south by more than a few percentage points
  • Auto-exit if a better opportunity presents itself
  • Keep an eye out for opportunities to buy residential real estate and related assets (like homebuilders, landlords, and residential REITs) when the asset's fundamentals look better than the average for the market and its price is below its 3-month moving average. Exit these positions once the price has jumped more than 10% in a day, risen by at least 20% overall, dips more than 3% below acquisition price, or 6 months have passed, whichever comes first.
  • Keep an eye out for opportunities to sell commercial real estate and related assets whose debt levels look particularly ugly and prices are above their 3-month moving averages. Exit these positions once the price has fallen more than 10% in a day, fallen by at least 20% overall, drifted up by 5% above acquisition price, or 3 months have passed, whichever comes first.
  • Ceterus paribus, slowly transition from US, dollar-denominated assets to those from China and other emerging markets. When the dollar temporarily spikes against these currencies, exit some laggard dollar-denominated assets and move the money into attractive Chinese and other emerging market assets. Avoid companies that export a lot to the US
  • Ceterus paribus, watch for opportunities to slowly move from investments in companies with non-dollar-denominated debt into those of similar companies with more dollar debt
  • If a headline turns up for one of the assets in your current portfolio, analyze the price vector (direction of change) after the headline and buy more or close out in order to beat everyone else to the punch.
  • Always maintain an appropriate level of rate hedge (for example, shorting treasuries, selling dollar futures/options, etc) in proportion to overall exposure, especially around significant announcements about the economy or the FOMC. If rates surge over a decent period of time, alert you to re-assess the strategy
  • And, of course, the security blankey: tell you if the model seems wrong compared with other peoples' behavior. Before accepting new rules from you, back-test them over some previous, similar time period to tell you you're being dumb. Oh, and slam on the brakes if the portfolio starts hemorrhaging money because of a flawed rule.
And Bob's your uncle! Ten years forward, investing like this will be commonplace. Believe me: I can see the future!

Illustrations Copyright

No comments: