Colin Twiggs Trading Diary
Layers Of Complexity
By Colin Twiggs
November 29, 1:00 a.m. ET (4:00 p.m. AET)
These extracts from my trading diary are for educational purposes and should not be interpreted as investment or trading advice. Full terms and conditions can be found at Terms of Use.
What do a nuclear power plant, an aircraft, a ship at sea, and financial markets have in common? They are all complex systems with layers of safety measures intended to protect the system from catastrophic failure. Richard Bookstaber, a Wall Street risk management specialist, points out that the very safety measures that are intended to protect the system add to the complexity and make unforeseen failures more likely.
There is another common feature: all the above systems are tightly coupled. Failure of one component or safety measure can impact on other components or measures in an unforeseen manner — with catastrophic results. This is where complexity bedevils the process, hiding points of failure from sight until the system is out of control.
A simple example is the fuel gauge on an aircraft. A much-needed safety measure, the gauge warns the pilot when his fuel is running low. But if the pilot comes to rely on the fuel gauge to tell whether his tanks are full at take-off, a faulty gauge could lead to disaster. The safety measure, the fuel gauge, adds a layer of complexity that may hide a half-empty fuel tank until the pilot is over the Atlantic. Multiply this simple example by thousands of times and you will have an idea of the complexity of any of the above systems.
USS Forrestal
The deck of an aircraft carrier at launch time is a highly complex system with tightly coupled events. Aircraft are raised from the hangar below on giant lifts, to the flight deck where they are fueled and armed before being launched from a steam-powered catapult. Witnessing such an event gives an insight into the months of drills and planned safety measures that enable the carrier to function as a well-organized system. There are few gridlocks and arguments about procedure. Everyone knows their job. Layers of safety measures are in place to prevent failure of one component in the system from impacting (domino-like) on others in the process.
Unfortunately, the complexity of the system increases the likelihood of unforeseen events. The USS Forrestal is a classic example. Operating in the Gulf of Tonkin during the Vietnam war, the carrier experienced a massive fire on its flight deck, claiming the lives of 134 crewmen and injuring 161 others. The worst fire in post WWII US naval history. In a surprising twist of fate, one of the few who witnessed the disaster and escape unscathed was a young navy pilot, LCDR John McCain.
Post-fire investigations revealed an unlikely sequence of events precipitated by reliance on some of the very safety measures intended to prevent such an occurrence. An electrical surge, when switching from standby to internal power, triggered a 5-inch rocket mounted in a pod under the wing of an F4 Phantom jet being prepared for take-off. The rockets were not supposed to be armed until later in the launch sequence. Large red tags added to the electrical safety pins to make them more visible, however, had caused the pin to become dislodged in high winds on the carrier deck. That in itself could not have caused the disaster. There was a backup safety measure requiring that the "pigtail" electrical connection to the rocket pod only be connected when the jet was mounted on the catapult, ready for launch. But launch delays caused by faulty connections had caused the weapons safety board to agree that "pigtails" be connected earlier, while the aircraft were queued on the flight deck. They were unaware of the potential danger from an electrical surge on transfer to internal power — because safety procedures had until then masked this potential threat.
A third safety measure, preventing rockets from detonating at close range, also failed to avert the disaster. The rocket struck the fuel tank on one of two Skyhawks being fueled on the opposite side of the flight deck, one of which was piloted by LCDR McCain. It failed to detonate, as designed, but punctured the fuel tank igniting the contents. McCain and several other pilots managed to scramble from their cockpits and jump to safety.
Fire crews, on standby for just such an occurrence, moved to extinguish the blaze before it could spread. But events had conspired against them. Two 1000-pound bombs had been dislodged from the Skyhawk and lay on the burning deck. The crew were trained for just such an eventuality and knew that the ordnance was designed to withstand intense heat for at least 2.5 minutes, giving them sufficient time to quell the blaze. This is where their luck ran out. Due to the particularly heavy bombing campaign the Navy was running short of ordnance and had switched to Korean War vintage bombs that lacked the same heat resistant poperties. One of the bombs detonated, wiping out the fire crews and killing one of the pilots. This set off a further chaotic chain of fires and detonations which caused the massive loss of life and nearly sank the ship.
Financial Markets
"What does this have to do with financial markets?" you may ask. Bookstaber, in his book A Demon Of Our Own Design, illustrates how complex trading strategies employed by banks and broker dealers, hedged using non-standard financial derivatives traded in over-the-counter markets, can have unforeseen consequences similar to disasters in other complex, tightly-coupled systems. An institution's need to liquidate its position in one financial instrument may impact not only on that market, but on other financial instruments as well. Attempts to extricate from complex trading positions can cause liquidity to dry up, resulting in a price fall sufficient to provoke further assets sales — in order to meet capital or liquidity requirements. Setting off a downward spiral. Counterparties aware of the institution's position may withdraw, not only from the market for particular instrument, but from other markets where the institution has known exposure. Either to take advantage of resultant lower prices or through fear of a downward spiral. This merely reinforces the liquidity squeeze.
Bookstaber's solution is simple. You cannot make complex systems safe by increased regulation or additional safety measures. These merely elevate the risk of unforeseen events. The only way to make complex systems safe is by simplifying them. In financial markets that means standardizing derivative products and trading them through a formal exchange. It also means restricting leverage in order to limit the impact of individual failure on the whole market. In addition, we can simplify the structure of complex institutions and require stronger capital-asset ratios to protect against failure. I also see no reason why standards should not be higher for large institutions in the too-big-to-fail category.
The Economy
The global economy is probably the most complex of all systems. It is also highly coupled, with failures in one area impacting in unforeseen ways on other areas. Adding safeguards and increasing regulation will not protect us from failure. It merely increases the complexity, making failure more likely. The only viable solution is to remove complexity by standardizing and simplifying processes. And minimizing interference with natural market forces.
The science of economic engineering is highly overrated. If similar safety standards were applied to other complex systems such as aircraft design, engineers would not even bother to fit landing gear. Because every engineered take-off inevitably leads to a crash.
Gold
Spot gold retraced to test the new support level at $800. Respect would signal a rally to the upper border of the descending broadening wedge formation, while failure would indicate another test of $700. The primary trend is down, and breakout below 700 would offer a target of $550, the June 2006 low.
Baltic Dry Index
The Baltic Dry Index reflects international trade, especially bulk commodity shipments of iron ore and coal. The index commenced another down-swing after a brief consolidation, warning that commodity exports from Australia, South Africa and Brazil, will continue to fall.
Interbank Lending
Yields on 3-month Treasury Bills are close to zero, signaling that uncertainty in financial markets remains high. Investors concerned about default risk prefer safety over yield.
USA
Dow Jones Industrial Average
The Dow rally appears to be running out of steam, though declining volume may be partly attributable to Thanksgiving weekend. Breakout above medium-term resistance at 9000 would signal another test of 10000. Reversal below 8000 is more likely, however, and would warn of another down-swing.
Long Term: The primary trend is down and penetration of the 2002 low of 7300 would offer a long-term target of 6000. This is calculated as 8000 - ( 10000 - 8000 ). Twiggs Money Flow (13-Week) shows a bullish divergence, but is yet to be confirmed by breakout from the downward trend channel. Recovery above 10000 is unlikely in the present climate.
Comments
Your author, Zak, made a fundamentally false statement - slipped in the middle at a critical point where it made a big difference to his argument, but where nobody may have noticed. That false statement is:
"You cannot make complex systems safe by increased regulation or additional safety measures. These merely elevate the risk of unforeseen events."
That is pragmatically and theoretically and demonstrably false.
For example, hark back to Jay Forrester (nobel laureate and economic genius) who applied Ross Ashby's cybernetic ideas to world economic systems. Fundamental to the whole science is the contrast between positive feedback systems and negative feedback systems. The former tend to spin out of control (while providing rewards to some in the meantime) but the latter tend to sputter in controlled fashion back to equilibrium.
The writer you quote here seems to be suggesting that there are ONLY positive feedback sorts of regulators. That is starkly false.
If you are anything like me, you don't have a lot of money to plop into safe or risky growth schemes. Okay, so the less we have the more we value what we do have. So how well do you trust the market in this environment. Would you take a few paychecks and drop them into a safe market fund today? I wouldn't. I might go find some collectable gold to invest in if I were to invest in anything long term right now. I would more likely go get a loan for some land right now, but that is pretty risky too. The difference is, I'm familiar with making land profitable.
jb
Very good question. Sorry for the delay in response to it. Am driving (again) across the country.
One of the great stories about good management and good science is the one about how the rivers and lakes in the industrial east were cleaned up. When it happened well, it was when someone applied the very general regulatory and cybernetic principle that a factory should TAKE IN its water from BELOW the place where they DUMPED their waste.