[Lexington]

This is a hangover post (I took the day off knowing the hangover would happen...anyway...)

The title is sarcastic - the people who think derivatives are evil are people who don't understand them. "Derivatives" refers to a wide range of instruments, some of which are very simple and easy to understand and hard to abuse, and some which are very complex and/or easy to abuse. I think the misunderstanding leads to unfortunate threads like this.

There are 2 1/2 basic kinds of derivatives:

Forward contracts are agreements to buy/sell something in the future. For example, if I want oil three months from now, I can buy a contract from someone and they will sell me oil three months from now - at an agreed price.

One benefit: if I know I'll need oil in three months, rather than worry about the price shooting up to $600, I can buy this contract now, and I know exactly how much I'll have to pay in three months. Likewise, if I want to sell corn in three months, and I'm worried about the market crashing to $0.10, I can get someone to agree to buy it from me for a current, reasonable, agreed upon price.

Now, is that really evil or mysterious in any way? I don't think so, but when people talk about derivatives, they're including in that word things as basic as forwards.

Options are agreements where the buyer has the right to buy/sell something in the future. Unlike a forward, I don't have to buy oil from you three months from now at the price we agreed on - if the market price is much lower then, I'd rather buy it on the open market and I won't use my option.

The benefit: Instead of being stuck paying/receiving a price that may be much higher or lower than what I can actually get on the market, I can choose not to exercise the option (exercising means, actually choosing to buy/sell at the price you wanted originally).

Options are far more fascinating (in my opinion) than forwards since they lead to cool mathematical shit like Black-Scholes

Now, I don't think that's very complicated or evil either (but then again, options are my specialty).

Swaps are agreements where someone agrees to pay/receive several times for something. Most swaps are interest rate swaps (in fact, the majority of all derivatives by notional value are interest rate swaps - even though most people have never heard of them...maybe people should) but interest rates are boring and...easy to understand but long to explain. I could get into LIBOR and eurodollars and crap like that but I'm yawning already just saying LIBOR.

So I'll give an example (albeit rarely used): say I need oil every three months. I can get a swap on oil - every three months I can pay for oil for a fixed price. Sometimes I'll win (the market price may be much higher) and sometimes I'll lose (I would rather buy it much cheaper on the market then) - hence "swap".

This is the 1/2 in the 2 1/2 kinds of derivatives: a swap can be built as a series of forward contracts. Instead of having this swap, I could have a forward contract buying oil 3 months from now, another buying oil 6 months from now, and so on, and it would do the same thing. But by convention, swaps are thought of as special and different from forwards (except to arbitrageurs and this aforementioned travesty of a thread)

The jargon file on derivatives

The examples used above were all about oil. This is called the underlying of the derivative. Derivatives are all about being derived from something, the something is the underlying.

The notional value of a derivative is not how much it's worth or how dangerous it is. The notional value is what it is based on - for instance, the notional value of the forward buying crude oil for $100 is $100 - the gain/loss of the contract is whatever the difference is between the price of oil and $100. For an interest rate swap, the notional value is extremely different from its actual risk and value. An interest rate swap on $10 billion in notional value (that is, $10 billion in bonds more or less) will probably only lead to a fraction of a percent of that in actual money changing hands due to changing interest rates. Potentially, yes, it could be $10 billion changing hands (it could be also much more if rates go the other way in the absurd extreme) but in all normal circumstances the actual value is tiny compared to the notional value for IRS.

While not going too far into overall finance, it's important to understand some concepts. One is asset class. Asset classes are categories of financial stuff: "equity" is an asset class that includes stocks, "fixed income" is an asset class that includes bonds, loans, and mortgages, "commodities" includes things like crude oil, corn, and iron.

The above examples of derivatives were examples of derivatives on commodities, because I think derivatives on commodities are clear to understand. One can just as easily imagine forwards, options, and swaps where the underlying is IBM (equity) or US treasury bonds (fixed income).

Derivatives: the evil/mysterious parts

The reason I mentioned asset class is because there are some kinds of derivatives which are special to fixed income assets (bonds/loans/mortgages) that have special features. "Special features" to a programmer means "a bug".

Credit default swaps (normally called just CDS) look like swaps. A CDS is a form of insurance - every couple months you pay something, and if the thing you have insurance on goes bankrupt, you get a huge payout.

Example: If I sold a CDS on Lehman Brothers in 2008, the buyer would pay me every couple months, then when Lehman went bankrupt, I would have to pay a ton (like normal insurance; the actual payout is buying at par the company's bonds even though they're almost worthless, so it almost has optionality features).

Although this looks like the crude oil swap I mentioned above (periodic payments/receipts) the "almost always payer" feature combined with the "massive potential receiver" feature is totally unlike most swaps. This feature - where a CDS seller can potentially lose enormous amounts of money rapidly if shit goes bad - is practically unique to CDS in Planet Swap - normal swaps don't usually approach their notional amounts nearly in actual cash moving around. CDS actually has payouts approaching the notional value of the swap unlike most swaps.

This is why AIG, an insurance company, went bankrupt: they sold CDS as insurance, and...faced an enormous fucking payout that destroyed them.

CDO/CLO/CMO/MBS/Mortgage tranches - These are often thought of as derivatives even though they don't fit into the forward/option/swap paradigm above. These are all fixed income securities derived from the most basic financial instrument around: the loan.

To a bank, a mortgage (or a loan) is a piece of paper that craps money out every month (when you pay your mortgage). This is extremely lucrative (pieces of paper that crap money out every month tend to be) but even more so when you can sell that money-crapping piece of paper to someone else.

But the other banks don't want to buy an individual mortgage because then they have to know who the money-crapping home-buyer is and that takes time and time is money. So instead, the bank (the structure is more complicated) takes a pile of mortgages and "pools" them - instead of Local Bank #1 selling J.P. Nobody's mortgage, they take, say, a thousand mortgages and put it together in a single bond - the holder gets a slice of whatever those thousand mortgages pay. That way the bank doesn't have to investigate every single mortgage-buyer, they just have to understand the overall mortgage pool.

This pool has unusual features: the mortgage pool has a fixed rate of return if the mortgage holders are solvent. If the mortgage holders go under, the banks foreclose on their houses (and the interest stops coming in).

Banks came up with an idea called tranches, which I think is just French for slices: take a mortgage pool that returns 5% a year (normally) and divide those returns up into separate instruments: one guy gets 2% first, then the next guy gets everything left. Normally, The Next Guy will do very well, but if the pool crashes (everyone starts foreclosing) then The Next Guy might be left with nothing because The One Guy gets his 2% first.

They have complicated terms for this like "equity" and "mezzanine" tranches but the idea is "someone pays less to get fucked first".

There are mathematical ways of modeling the probability of default (bankruptcy) and correlations between bankruptcies (copula). These were fundamental to the valuation of these securities but failed to understand rising correlations and tail risk. In my opinion - but I'm just a history major.

The evil part of this is that it's very easy (in 2008 at least) to take a bunch of mortgages, tranche it like this, and sell off the top few tranches and say "It's as good as treasury bonds!" But these tranches are not very transparent - and when suddenly the market dived...the upper tranches crashed beyond all expectations.

That last paragraph explains the financial crisis in its totality.

[mikema63]

This should be stickied.

[Davea8]

Well, I'm sure you successfully put across the point to all present that you are smart. But are you saying that such instruments had no detrimental effect that played no part in the housing crash that continues, or in any other area of the economy? ... Just the last paragraph covers all the "evil"?

[Soixante-Retard]

"People fear what they don't understand and hate what they can't conquer."

[Rei Murasame]

For the record, even though I started the thread that Lexington links to as the 'travesty of a thread', I don't actually have a problem with derivatives, even though the article in the OP and many of the other posters in the thread seem to. I only have a problem with power structures which incentivise the misuse of derivatives.

[Rainbow Crow]

I hope you're not suggesting that we actually study issues before we start posting about them, Lex. That would be a questioning of the whole Pofoific method.

[Baff]

Well, I'm sure you successfully put across the point to all present that you are smart. But are you saying that such instruments had no detrimental effect that played no part in the housing crash that continues, or in any other area of the economy? ... Just the last paragraph covers all the "evil"?

Knowledgable Dave. Not smart. Educated. Wise.
Possesing experience of the subject matter. And thats the reason I come here. To find people who have actual experience of the subject matter.
Opinion or intellect doesn't cut the mustard. Either you know what you are taklking about or you don't.


Those who call banking evil have no desire to understand. No desire for wisdom.
Only the desire to hang a convenient scapegoat.

It was always going to be either the politicians or the bankers.

[TropicalK]

Derivatives are not evil in themselves, but there are problems associated with them.

Banks do not hold enough collateral to back CDS, and using customer accounts as collateral should be banned completely.
The asymmetric payoffs of far-out of the money options cause bank traders to gamble excessively, knowing that they have capped losses.
I know a person at my previous employer who earned a million a month, and then dropped $400 million in a day. He retired rich.

The problem with mortgage tranches is that the major banks bribed rating agencies to grade them AAA. The amount of fraud committed essentially makes them lemons markets. Thus uninvestable if you do not have inside information like John Paulson and his deals with Goldman Sachs.

Forward markets and swaps are largely fine. I'm actually more concerned with corporate bond markets where being traded through is the modus operandi with no legitimate central electronic exchange.

If you are a high net worth individual receiving calls from Goldman Sachs or JP Morgan, then you are a muppet imbecile if you follow their trade recommendations. The banks massive trading profits don't just materialize, they come from the preyed upon sheeple.

In summation, the derivatives are not the problem. The problem is bankers and traders abusing asymmetric payoffs, withholding trades and prices from the market in order to con unsavy investors and cause lit markets to have bustic behavior at the bid-ask level.

[Baff]

Banks do not hold enough collateral to back CDS, and using customer accounts as collateral should be banned completely.

Disagree.
Money that is not being used is an opportunity wasted.

Whether my savings are invested in a savings account, an insurance scheme or government bonds or stocks and shares... makes no difference to me. Only the return does.

I fail to see why a savings account is somehow a sacred investment and the others not.


The problem with CDS is simple. Someone defaulted.
The insurers had to pay out.

Now we could say that perhaps the rate of return on certain CDS's was poorly/irresponsably calculated in the first place, or equally that over a longer period of time they may turn out to have been correct. Insurance is after all a long game and certainly not a precise science.
If I was to give a more common example such as car insurance, it is possible that on my very first day of getting insured that I write off my car. The insurance company would then be making an immediate loss on my policy. That does not mean however that they set a bad/irresponsable rate of insurance for me, only that the precise timing and scale of any repayment is not 100% predictable. Just a best guess.

It should also be noted that by and large, the banks/insurers did have enough capital to cover their CDS.

[Kaiserschmarrn]

I'm not sure that derivatives themselves are so poorly understood. After all, your average person doesn't have to fully understand the maths behind them, but is perfectly able to grasp the nature of an option, forward, or swap contract. What is obscure (to some extent) is the workings of the financial markets and how they facilitated bringing giants like AIG to their knees. And this happened while it was thought that "risk was more evenly spread throughout the markets than ever", apparently giving the majority of market participants a false sense of security.

Considering this, it's not surprising that there is great suspicion now. If this was a genuine believe among financial institutions and regulators, then one has to question their ability to understand the markets themselves. And since this is (at the very least) a very uncomfortable thought, it's easy to jump to the conclusion that the system is completely corrupt and dismiss any explanation attempts by people who work in the industry. On the other hand, it doesn't help that many knowledgeable people are supremely arrogant (I'm not referring to you, personally, Lexington) and lightly dismiss concerns while implying that everybody else is just an ignoramus

[Nunt]

Excellent post.

[ralfy]

About Paul Wilmott:

http://www.cqf.com/faculty/dr-paul-wilmott

"Big Risk: $1.2 Quadrillion Derivatives Market Dwarfs World GDP"

http://www.dailyfinance.com/2010/06/09/ ... arket-gdp/

"Top Derivatives Expert Estimates Size of the Global Derivatives Market at $1,200 Trillion Dollars … 20 Times Larger than the Global Economy"

http://www.washingtonsblog.com/2012/05/ ... arket.html

"Revenge of the Nerd"

http://www.thedailybeast.com/newsweek/2 ... -nerd.html

From Wilmott:

"The Problem with Derivatives, Quants, and Risk Management Today"

http://www.qfinance.com/capital-markets ... day?page=1

"Q&A with Quant Expert Paul Wilmott: The Fundamental Flaw of Nearly All Risk Models"

http://www.advancedtrading.com/quant/qa ... /231901539

[liberallad]

I think the mystery behind derivatives like credit default swaps speak to a larger lack of information and understanding in how financial markets work. Particularly, financial regulations (Glass-Steagall, etc.) have altered markets and created inexplicable loopholes that, as far as I know, few beyond the insular world of front office banking, politics, and financial journalism know the technicalities of.

[oppose_obama]

best thread ive read/.

[wat0n]

This is an awesome thread. It should be stickied.

[Eran]

I share others' appreciation for the OP.

As a finance professional, I would make two comments:
1. While most derivatives users are professionals who understand their behaviour fairly well, some complex derivatives have been sold (packaged as structured notes) into the retail market, where they were purchased by "orphans and widows" who may well have misunderstood how bad a deal they were handed. I worked on the Equity Desk of Lehman Brothers in 2006-2007, and structuring attractive-looking retail products was a lucrative, if semi-fraudulent business.

2. The financial crisis wasn't caused by misunderstood derivatives, but by misunderstood markets. Specifically, risk managers (myself included) routinely use the recent historic record as an indicator of the range of plausible market moves. When markets (including, but not limited to the housing market) move in unprecedented manner, risk models break down. Derivatives often magnify the impact of such unprecedented moves.

[oppose_obama]

Can you give some examples of complex instruments?

[Lexington]

What I wanted to have was a post I could refer to when people said DERIVATIVES IZ EVUL! and explained that...well, they're not talking about derivatives, whatever they're talking about is a part of derivatives.

But are you saying that such instruments had no detrimental effect that played no part in the housing crash that continues, or in any other area of the economy? ... Just the last paragraph covers all the "evil"?

My view is that no derivative is evil in itself - it's the way it's regulated, understood, and rated that can cause it to be used in evil ways. Even a CDS, which is usually brought up as THE MOST EVIL OF ALL THINGS IN FINANCE is fundamentally sensible - it's insurance. It's a very good idea in concept - if you can disperse the risk that a company goes bankrupt across the financial system, you reduce the risk that it can destroy you (and, after you, other people). In principle, that's very very good.

As TropicalK rightly points out, we can make that quite safe (by margining and collateral and so on).

I'm not sure that derivatives themselves are so poorly understood. After all, your average person doesn't have to fully understand the maths behind them, but is perfectly able to grasp the nature of an option, forward, or swap contract.

That was the point of the post...I don't expect most people to understand Black-Scholes or local volatility. People don't need to understand these things to analyze the role of derivatives any more than they need to understand day count conventions to understand bonds conceptually. Yet people decry derivatives as evil often enough when they're really echoing some rational complaint (often many times echoed) about some particular categories of derivatives...while not understanding what a forward is. The common understanding seems to be: forward? CDS? option? is an option even a derivative!? Derivatives are evil!

On the other hand, it doesn't help that many knowledgeable people are supremely arrogant (I'm not referring to you, personally, Lexington) and lightly dismiss concerns while implying that everybody else is just an ignoramus

I don't mean to dismiss concerns about derivatives (in fact I mentioned CDS and MBS tranching specifically...I should have gone on longer, but again I was hungover and passed out around the time I pressed the post button a week ago)

Coincidentally, there was an article very much more and more sinisterly about this (about the fate of Dodd-Frank in the US) just recently:

Another swinging mace in this stage of the rule-making gauntlet is what Kelleher, the head of Better Markets, calls the “Wall Street Fog Machine.” “They come at you with this jargon,” he said. “They want to make you feel like it’s too complicated for you to understand. You’re stupid, and they’re the only ones who get it—that’s the end game.” This is particularly true when it comes to financial products, like customized swaps, which traders on Wall Street have spent the last decade designing precisely in order to swindle their clients.

It is my honest intention in this thread to explain what derivatives are. (although 2/3rds of the swaps they're talking about I likely don't understand since...I rarely work with swaps!)

If I am unclear about any point in this thread (and it may be just because I don't understand it either) it is not because I'm trying to hide something under the kimono.

http://www.cqf.com/faculty/dr-paul-wilmott

"Big Risk: $1.2 Quadrillion Derivatives Market Dwarfs World GDP"

http://www.dailyfinance.com/2010/06/09/ ... arket-gdp/

"Top Derivatives Expert Estimates Size of the Global Derivatives Market at $1,200 Trillion Dollars … 20 Times Larger than the Global Economy"

This was exactly why I mentioned notional value:

The notional value of a derivative is not how much it's worth or how dangerous it is. The notional value is what it is based on - for instance, the notional value of the forward buying crude oil for $100 is $100 - the gain/loss of the contract is whatever the difference is between the price of oil and $100. For an interest rate swap, the notional value is extremely different from its actual risk and value. An interest rate swap on $10 billion in notional value (that is, $10 billion in bonds more or less) will probably only lead to a fraction of a percent of that in actual money changing hands due to changing interest rates. Potentially, yes, it could be $10 billion changing hands (it could be also much more if rates go the other way in the absurd extreme) but in all normal circumstances the actual value is tiny compared to the notional value for IRS.

That "quadrillion" number is notional value. If I say I'm going to buy a bag of chips from you tomorrow that's a $1 notional value but in all likelihood I'll be winning or losing about $0.01 tomorrow...but in notional value the derivatives in the planet increased by $1 anyway.

I am not precipitating armageddon in trading a cent with you tomorrow, really?

Extrapolate to 3 months or 2 years time.

Can you give some examples of complex instruments?

I tried to protect you!

But if you have to look into structured notes, look on the eyes of Cthulu: RBC's "Return Optimization Securities with Partial Protection" note...

(of the top of google)

I'll try translating later, right now it's party time.

[ralfy]

This was exactly why I mentioned notional value:

"The notional value of a derivative is not how much it's worth or how dangerous it is. The notional value is what it is based on - for instance, the notional value of the forward buying crude oil for $100 is $100 - the gain/loss of the contract is whatever the difference is between the price of oil and $100. For an interest rate swap, the notional value is extremely different from its actual risk and value. An interest rate swap on $10 billion in notional value (that is, $10 billion in bonds more or less) will probably only lead to a fraction of a percent of that in actual money changing hands due to changing interest rates. Potentially, yes, it could be $10 billion changing hands (it could be also much more if rates go the other way in the absurd extreme) but in all normal circumstances the actual value is tiny compared to the notional value for IRS."

That "quadrillion" number is notional value. If I say I'm going to buy a bag of chips from you tomorrow that's a $1 notional value but in all likelihood I'll be winning or losing about $0.01 tomorrow...but in notional value the derivatives in the planet increased by $1 anyway.

As explained to you repeatedly in another thread, if only the world operated through "normal circumstances," but it doesn't.

Hence, Wilmott's points.

[TropicalK]

Lexington,

You are setting your opponent's argument to such a weak case that its borderline straw-manning.

Derivatives inherently are contractual concepts, with no inherent goal, reason, objective, or life-force. Thus they cannot be "evil" as they are just pieces of paper or electrons. I do not believe anything is evil, but that's another topic.

Scientifically, if we were to set derivatives are not evil as the null hypothesis, what sort of evidence would one need to reject it?

It is far more sensible to critique derivatives by how the are used and abused by others than by some mystical inherent "evilness."