I’ve noticed over the years of following business and the trading of financial instruments, that there is an obvious pattern followed in virtually every area of study.
It should come as no surprise that there is an increase in the fragility of every system over any period of time when there is relative stability. This pattern seems to exist in most every area of study I’ve come across. The fragility will increase at least until it matches the current level of stability of the environment, and then become exactly as fragile as that environment will allow. It will also become more fragile over time such that eventually even the smallest increased deviation from the stability will wreak havok on the participants in the system.
There are 2 required components for an environment to follow this pattern:
- The environment must have a probability of a volatile event occurring that is not a common occurrence and that does not come in predictable intervals
- There must be an advantage for participants in becoming more efficient should the event not occur
When these two requirements are met, all it then takes is time. Naturally, there will be a randomly occurring period with no event occurring that exceeds either the memory of the participants or the usual frequency of the event, and the fragility cycle begins. The longer the stability lasts, the more destructive it will be when it occurs.
An example is supply chains. When times are tough and unpredictable, supply chains are robust and have excess capacity to compensate for uncertainty. When times become peaceful and (apparently) predictable, supply chains become lean and exacting, with little warehouse space needed as shocks to the system become less frequent. After years of this, they become so efficient, they will often be “just in time” and have little excess capacity or any sort of buffer in case of disaster. They become so efficient and fragile that it can move from an inventory that can last last months without resupply, to a handful of days or fewer. In theory, with perfect efficiency, and also perfect fragility, there is no inventory and the goods arrive at the moment they are sold.
This makes sense to me because in a peaceful yet competitive environment, with short term payouts, the excess capacity within the supply chain and the excess inventory sitting in the warehouse lower profitability. Efficiency and lean systems win out competitively in stable times.
The same thing happens in evolution. When a species is evolving without much change to its environment, it will evolve more and more specialized to that environment. For example, birds or plants on an isolated island near the equator would be expected to be less evolved towards change than those with natural predators and changing seasons. Over time their specialization and fragility will both increase such that they can make the most of their unchanging environment. This fragility should increase over time until the species become “perfectly” evolved in their environment and utterly unable to survive in any other. Eventually, with almost any change to this environment (the addition of a predator for example, or a change in climate, or the loss of a single food source), the evolved fragile species will go extinct as it cannot adapt quickly enough and has become too specialized. I wonder if some major historical extinction events are more likely after disproportionately long periods of stability in the environment.
The impact of this should be as follows: if we have a system with unpredictable volatility and chaos, with large intervals of peace in between, we should expect many systems that become highly specialized during those peaceful periods will fail when volatility arises. Additionally, those systems that are adapted to a volatile environment and are unable to become more efficient to take advantage of an unchanging one, should gradually fail over time against those leaner and more specialized systems once peaceful times return. The longer the peace time lasts, the more efficient and the more fragile everything related to it becomes, until even a tiny shock will destroy the system.
We see this happen also with bid/offer spreads and stock volatility. When volatility declines, bid/offer spreads decline, as market makers need to be competitive with their spreads, but not so competitive that volatility runs them over. In lower volatility (peaceful) times, the tight spread will continue to make money and most of the opportunity in wider spreads will be missed. They start taking on risk that will lose should volatility increase. This will work for them up to the very moment there is a volatility spike. When volatility suddenly increases, we will see the market makers with the tight spreads get run over and lose consistently, until spreads are widened. Many will go out of business, as their business model evolved to a market condition that no longer exists.
Imagine you are able to sell lottery tickets, and each ticket has an independent, 1 in 1 million chance, of winning a million dollars. Each ticket should have a fair value of $1. Say you sell 1 ticket a day for $1 each, and are required to pay should any ticket hit the lottery, paying $1 million for every ticket that hits the lottery. Day after day, you sell your one ticket and collect $1, and nobody hits the jackpot. You’re feeling pretty good because the likelihood anyone hits the jackpot is incredibly slim on any given day. You don’t even bother to save the money, but you just spend it as years go by and nobody hits the jackpot. Eventually, you start selling them for less than a dollar and start selling more tickets. Day after day, you collect money and feel no risk. Even though your expected value is now negative, since fair value is $1 and you are selling the tickets for less than that, you still appear to be making money day after day as long as nobody hits the jackpot. The chances of anyone hitting the jackpot are small. So far so good, until someone, or maybe multiple people, eventually hit the jackpot. At which point you lose a fortune and “blow up” and are completely unable to pay and are out of business. This analogy is what is happening when environments are peaceful with regular and unpredictable future shocks that occur with rare frequency. These are by far the most common environments you will come across.
Financial markets themselves become fragile. Once again, over time, and without any crashes, they become more efficient and leveraged. Eventually participants become so leveraged that even a small drop will wipe them out and cascade into a major crash. When housing goes up year after year for 20 years, people start believing it will remain that way forever. They leverage themselves more and more, putting down less money and speculating at higher priced purchases, until even the smallest drop in the market will cascade into defaults. The likelihood of a spectacular crash increases dramatically.
The idea that “this time is different” exists because of periods of peaceful environments that convince people they will last forever. Because the timing of the upheaval is rarely predictable, an event that usually happens once every 10 years, may happen twice in a single year and then go 50 years without occurring, causing spectacular destruction due to its irregularity and the long peaceful time where everything becomes fragile to the eventual shock.
We can also see this happen with governments. The longer the period without war or social upheaval, the less capable governments are to handle extreme events. If you go years without war, the quality of your standing army declines as the government becomes more efficient by spending money elsewhere and reducing the budget of the standing army, essentially becoming more fragile. Eventually the people of the nation feel any spending on an army is wasted money. They will believe war is a thing of the past and wipe out their preparedness. We can see an example of this in how most of European nations currently spend less as a percentage of GDP on national defense than at any time in history, while existing under the Pax Americana. Do they think war will not come again? Do they think that their borders will never again be at risk of invasion? Perhaps they are right, but if history is any indication, war will likely at some unknown future time, come again to their borders and they will not be prepared for it due to decades of peace.
Think of hurricane/flood insurance. If hurricanes are random yet somewhat regular events, you would expect insurance to remain with generally steady rates, all other factors being equal. But it doesn’t. Instead, insurance is highest when a hurricane has just taken place. It then drops down over time until it reaches its lowest level just prior to the next hurricane, after which it then skyrockets. This is because, during the peaceful period, the insurer and all competing insurers drop rates over time when nothing has happened, and as people expect lower rates when it appears to rarely flood in their region. The rates follow the demand and recency of the event. A homeowner in a flood zone where it hasn’t flooded for 20 years isn’t thinking a flood is likely, even if it happened approximately once every 15 years prior. A homeowner who just had their house destroyed by a flood is almost certain to pay the price for insurance even if the event is rare and has an equal percentage change of happening in a given year. And all of this thinking and pricing will occur without any change whatsoever to the actual frequency of flooding.
This evolutionary flaw, if it is indeed a flaw, has a terrible impact if there is an equal chance of something happening every year, and that chance is low. When a system becomes less volatile, the pace with which efficiency increases is slow and can take years to evolve, not immediately wiping out the less efficient and less fragile participants. Peace breaks out slowly. But war or other volatile events happen in an instant. There is no time for evolution or adjustment when movement in a system goes from peaceful to volatile. It destroys those not prepared in a flash.
It means every few years or decades, we will have an event that will completely wipe out an industry, a government, a financial institution, or a species. How do you convince someone to make less profit and be less efficient than their competitors, year after year, for an event that happens once every 10 years, or once every hundred or even 1000 years? If we knew we were unprepared, as a species, for a meteor strike that will happen as a 0.05% chance every year, we must not allow ourselves to become so fragile with our species over time that it makes us extinct.
How many of us still own a land line for a phone? Land lines are far more robust than our cell phone systems and create a redundancy that costs money every year. Most of us have dropped them by now, as if complete dependency on our cell phones makes sense.
How many of us, (people, governments, companies), are prepared for a once in 500 year solar? One the size of the flare in the 1800 would likely destroy all electronics on the planet. How many live in a major city and have less than a few days food supply in case of an unforeseen emergency?
I am as guilty as anyone, I have no landline and it doesn’t bother me. But it should.
We must maintain some excess capacity and inventory in the face of everyone who looks shorter term and sees an immediate benefit to being more efficient. How do you convince someone to stay prepared and pay that expense when the event is unlikely to happen in their lifetime? I don’t know the answer, but we must figure this out if we are to survive as a species. We are so effective in our economies that we become fragile and efficient faster than ever in history, and as a result, we are less prepared than ever before to handle a shock in any environment. Because wars, supply chain shocks, and negative environmental events have happened with less frequency lately, we are more interconnected than ever before and we risk a shock to one environment causing cascading shocks in all environments, creating a greater fragility than ever before as all systems become more dependent and correlated to one another in