Someone told me last week, re a proposed hedging strategy, that it wasn't hedging properly so-called, because it was 'merely transferring risk'. This set me thinking. What else would hedging be but 'transferring risk'. Isn't it always the case that we can only reduce risk to ourselves by transferring it to another party. Is there some sort of law, like the law of energy or momentum, that requires risk always to be conserved when individuals reduce their exposure to it?

I think there are fairly deep questions lurking here. Superficially, we can certainly create and eliminate risk out of thin air through financial contracting. If you start a poker game with relatively high stakes, it is hard to argue that you didn't just increase the amount of risk each player is exposed to. Conversely, by shutting it down, you make that particular source of risk go away. On the other hand, the risk of $10 billion worth of damage to LA in an earthquake cannot be eliminated through insurance or derivatives or the like, but it can be re-distributed thusly in ways that may be more economically efficient. You can reduce the earthquake damage risk through physical means, e.g. via adherence to stricter building codes. Unfortunately, in a complex economy it would seem that most sources of risk lie somewhere in between the purely synthetic (the poker game -- which, incidentally, people voluntarily participate in) and "acts of god" (earthquakes), which makes your question really hard to answer cleanly. I would love to hear Aaron's view on this, but he has been notably absent for a while.

The answer is mixed and depends on the local unit of analysis and volume of hedging:Imagine an investment bank with zero risk -- it's a stretch but give it a try!1. Imagine the bank creates and sells an OTC instrument X to buyer A. If the payoff terms of the instrument are a function of unknown environmental variables (e.g. the market price of wheat in 6 months), then buyer A is exposed to Risk(X) and the bank is exposed to Risk(not(X))*. 2. To mitigate the bank's new risk, the bank might hedge it's exposure by purchasing one or more instruments from seller B that replicate Risk(X) and thus cancel the bank's risk.* Now buyer A is exposed to Risk(X) and seller B is exposed to Risk(not(X)).3. Now what if buyer A and seller B are economically linked either directly or indirectly. Perhaps buyer A bakes bread, worries that the price of wheat might be high in the future, and bought a contract from the bank to fix the future price of wheat. Perhaps seller B is a farmer, worried that the price of wheat might be low in the future, and sold a contract to fix the future price of wheat. Whether A and B transact directly or each buys and sells wheat at market prices (and uses their derivative contract to offset the market price back to some fixed price*) does not effect the outcome (although it is hugely important to the friction-free functioning of the system).Overall, event #1 seems to have created risk in the system, event #2 transferred that risk, and event #3 showed that events #1 and #2 may have actually eliminated risk.* This first analysis ignores counterparty risk!To the extent that the parties in the system might fail or choose not to fulfill their obligation (or the parties might disagree on the meaning of the contract terms), the risks of these instruments and transactions might be different than the parties expect.If wheat prices surge (e.g., hyperinflation), then the bread baker is going to demand that the bank pay the difference between the prevailing ultra-high market price and the contract's fixed price. But if the bank has failed in the economic turmoil, then the contract buyer isn't going to be paid. Even if the bank hedged the risk, bought a contract from seller B, and seller B pays the bank, buyer A would be just another creditor to the failed bank and might get pennies on the dollar.And this is where the entire risk creation/transfer process becomes much more complex. At one level, all of these transactions might be for the purposes of mitigating risks: buyer A bought the contract to hedge their risk exposure to wheat, the bank bought a contract from seller B to hedge it's exposure to the first contract, and seller B sold a contract to the bank for purposes of hedging their risk exposure to wheat. At that level, these activities benefit everyone and reduce everyone's risk. Yet to the extent that the players in the system think they can eliminate risks by hedging (which really is transferring risk), they may take on greater and greater volumes of business (especially if interest rates are near-zero). If the volume of business becomes unsustainable or some other severe shock affects the system, the counterparties may start to fail causing cascading failures through the system (e.g., 2008!).At low volumes, hedging stabilizes the system. At high volumes, it may destabilize the system.

Nice summary, bearish. Other issues are the distinction between idiosyncratic and systematic risk.The small scale pokergame creates idiosyncratic risk out of thin air, but creates no systematic risk for society as a whole. Systematic risks can(sometimes) be shared by diversification or derivatives. The main problem with having a 'conservation' law seems to me that even systematic risks can be created out of thin air, equity markets being a good example, and the excesses leading to theFinancial Crisis being egregious ones. edit: T4A, sorry for the apparent immediate follow -- we were both typing at the same time.

Those are excellent points, bearish and Alan.The "acts of god" risks are examples of "fool me once, shame on you, fool me twice, shame on me" risks. The first settlers in LA did not know of the earthquake risk and were thus exposed to true "acts of god". Subsequent, post-quake settlers and real estate developers created synthetic risk. Thus, knowledge of a risk converts it from an "act of god" risk into a synthetic risk.Yet there is an insidious feedback loop in that system of knowledge of risk and synthesis of risk. In the case of floods, flood prone lands are devalued. Yet this makes the land attractive to price-sensitive land buyers (e.g., the poor). Thus, the people least likely to be able to afford flood risks are the ones most likely to buy flood-prone land. Obviously, one can mandate the purchase of flood (or earthquake) insurance but price sensitive buyers will decry the huge expense of such insurance and seek ways to avoid buying it.Whether an idiosyncratic risk becomes a systemic risk depends on the volume of the idiosyncratic risk, the coupling of idiosyncratic risk takers in the system, and the buffers in the nodes that couple those risk takers together.

QuoteOriginally posted by: Traden4AlphaThose are excellent points, bearish and Alan.The "acts of god" risks are examples of "fool me once, shame on you, fool me twice, shame on me" risks. The first settlers in LA did not know of the earthquake risk and were thus exposed to true "acts of god". Subsequent, post-quake settlers and real estate developers created synthetic risk. Thus, knowledge of a risk converts it from an "act of god" risk into a synthetic risk.Yet there is an insidious feedback loop in that system of knowledge of risk and synthesis of risk. In the case of floods, flood prone lands are devalued. Yet this makes the land attractive to price-sensitive land buyers (e.g., the poor). Thus, the people least likely to be able to afford flood risks are the ones most likely to buy flood-prone land. Obviously, one can mandate the purchase of flood (or earthquake) insurance but price sensitive buyers will dreary the huge expense of such insurance and seek ways to avoid buying it.Whether an idiosyncratic risk becomes a systemic risk depends on the volume of the idiosyncratic risk, the coupling of idiosyncratic risk takers in the system, and the buffers in the nodes that couple those risk takers together.This is reminiscent of lectures in the 90s at university were knowledge of Black-Scholes was slowing being to drift back to physics departments.A comment heard frequently at those lectures was physicists would be good at valuing derivatives as this was like many diffusionproblems they had solved in various different branches of physics. But they would stumble in making useful contributionsto economics because there were no conservation principles to lean back on to make analogies with physical systems. A couple of decades later that doesn't seem so far off the mark. The more one looks deeper for conservation principles, the more queasy one gets about economics, I find.

Portfolio insurance, 1987 crash. There are no laws. Nothing is conserved. Quant finance is more like math biology than physics.It's amazing that people still think like physicists. I blame the French. Use the tools of physics but don't be stupid and believe any of the resulting models.Surely everyone has known all of this for decades??? I guess not. I blame Masters in Finance programmes.P

QuoteOriginally posted by: PaulPortfolio insurance, 1987 crash. There are no laws. Nothing is conserved. Quant finance is more like math biology than physics.It's amazing that people still think like physicists. I blame the French. Use the tools of physics but don't be stupid and believe any of the resulting models.Surely everyone has known all of this for decades??? I guess not. I blame Masters in Finance programmes.PI just blame the french

QuoteOriginally posted by: daveangelQuoteOriginally posted by: PaulPortfolio insurance, 1987 crash. There are no laws. Nothing is conserved. Quant finance is more like math biology than physics.It's amazing that people still think like physicists. I blame the French. Use the tools of physics but don't be stupid and believe any of the resulting models.Surely everyone has known all of this for decades??? I guess not. I blame Masters in Finance programmes.PI just blame the frenchExcept Francoise Hardy and Johnny Halliday, what?

QuoteOriginally posted by: PaulThere are no laws. Nothing is conserved. Quant finance is more like math biology than physics.P"Soddy wrote that financial debts grew exponentially at compound interest but the real economy was based on exhaustible stocks of fossil fuels."And the recent housing and internet bubbles are good example, both based on whacky assumptions. It was more like a birth and death process. But that's not what clients want to hear.

QuoteOriginally posted by: CuchulainnQuoteOriginally posted by: PaulThere are no laws. Nothing is conserved. Quant finance is more like math biology than physics.P"Soddy wrote that financial debts grew exponentially at compound interest but the real economy was based on exhaustible stocks of fossil fuels."And the recent housing and internet bubbles are good example, both based on whacky assumptions. It was more like a birth and death process. But that's not what clients want to hear.Yes, it is much more like math biology than physics.In the short-run, some economic systems do seem to have conservation of something. Simple competition in the short run or simple trading in a market seems like a zero-sum game. But the long-term and broader picture is often a positive-sum game. There's people that squabble over the pie (and physics models for that squabble) and there's people that make the pie bigger (this part is not well modeled and is more "biological" in character). The issue with bubbles is that it's hard to know when the economy is in a pie-growing phase and when its just exuberant belief in the pie-growing phase that is creating the appearance of growth. Yet unlike the tulip bubble or south seas bubble, some bubbles do have long-term beneficial impacts. The internet bubble did convert billions of investor dollars into a huge boom in hardware and software. Amazon.com is now worth 6X it's internet bubble peak so perhaps the internet bubble has less in common with other bubbles than one might think. But Soddy is mostly wrong about fossil fuels. Although fossil fuels may be exhaustible (interestingly, the number of years of reserves seems to keep growing), the ways in which one can create economic value from a unit of fossil fuel is almost inexhaustible. In constant dollar terms, the US produces about 2.5X the GDP per barrel of oil as it did just prior to the OPEC energy crisis in the early 1970s. And I'd bet most of Europe is the same. The oil might be exhaustible but that does not imply that the economic output tied to that oil is exhaustible.The little part where Soddy was right is that the growth in system (both the human economic system and biological systems) is driven by an outside energy source. These systems are open systems and that means they don't have to obey all those ugly physical conservation and entropy laws. These systems convert energy into complexity and in the economic realm, that complexity is valuable. There probably are physical bounds on the efficiency of conversion of physical energy into economic value but I'd wager that we are a long way away from those bounds. And there are bounds on the amount of accessible energy but, in this case, it's easy to show that we are a long way from those in that we've only just begun to tap into the tiniest fraction of solar energy.

This is fascinating conversation, but I think it went on tangent to the original question... Pretty much everybody here says that overreaching concept of "risk" is not conserved/conservable (because it is not definable, quantifiable?), but the (idealized) finance is all about conservation of price-risk relationship.

QuoteOriginally posted by: And2This is fascinating conversation, but I think it went on tangent to the original question... Pretty much everybody here says that overreaching concept of "risk" is not conserved/conservable (because it is not definable, quantifiable?), but the (idealized) finance is all about conservation of price-risk relationship.I agree. Risk on its own means nothing.What is a concrete example of this "conservation of price-risk relationship". I think that's a good term you use.What about conservation of price-risk of a house (as asset etc.) How can it be modeled. It is a living things while physics only works with dead things like quarks. Math Biology is better as Paul is suggesting? Why not, the previous model was not a runaway success.

Risk is easy to define in terms of variations in outcomes, very easy to create through contingent contracts, and seemingly easy to destroy through hedging and diversification (which involve even more of these contingent contracts). The price-risk relationship cannot be conserved in liquid and complete markets because someone can always find a way to buy the risky asset, hedge off the bulk of the variation that seems to create the risk, and thus steal the risk premium that more simplistic buyers of the asset think they are getting. (Of course, markets are never as liquid and complete as this theory requires but the effect is there.)Our tangent concerned the deeper issue that the belief that one can hedge risk leads people to actions that create hidden (and unhedgable) risks. The failure of physics to model this arises from physic's core assumption in an invariant universe -- the rules are fixed and all we need to do is discover them through the clever application of the scientific method. In economic systems, people keep creating and modifying the rules. Worse, these rule creators often have incentives to intentionally hide the rules or capriciously change the rules for their own gain (where their gains are the result of other rules created by other people for other reasons).