I understand how positions can offset and cancel each other out. What I don't understand is if the notional value is never fully redeemed, why use leverage at all? That is, if only the spot prices are used to actually exchange money, why inflate the value of a trade with leverage?

I apologize if I'm misusing terms - I'm not a finance expert.

The only explanation I can think of is that leverage is used to exploit dumb money that isn't hedged correctly. In which case, leverage is a barrier to entry that diminishes the number of participants in a market and reinforces cartel behavior. Which I would think is a bad thing.

The notional is purely a reference amount. For the vast majority of derivatives the notional amount is never exchanged (cross-currency basis swaps being a notable exception).

E.g a simple interest rate swap: I agree to pay you 2% per annum on a reference notional of 100m for 10 years, and in exchange you pay me libor referencing the same notional. At the end of first year let's say libor fixes at 1.98%. Hence I pay you a net 0.02% (2% - 1.98%) of 100m, i.e 20,000. We do this every year for 10 years.

It is just a way to hedge interest rate exposure - e.g I have borrowed 100m paying a Libor-linked interest rate and don't want to run the risk that Libor spikes to 10% or whatever.

There's nothing complicated about this, at least in principle. Derivatives just allow people to hedge out exposures they don't want.

Consider that borrowing 10 year money at the 6-month LIBOR rate is equivalent to borrowing at LIBOR for 6 months and rolling the debt 19 times.

The swap market provides a linkage between the money market ( borrowing < 1 year) and the capital market (borrowing > 1 year), and allows one to borrow (or lend) where the comparative advantage is better.

The 10 year swap rate is then a market number which represents the relative demand (and supply) within the money & capital markets.

I see. So if you pay cash now you are vulnerable to interest rate swings. Whereas if you use leverage to spread the term of your investment out, you can average the interest rate across the business cycle and achieve more predictable behavior. Is that correct?

I think point of the example was not to say that you could use leverage. You're borrowing the same amount that you're lending.

The idea of the interest rate swap is simply that some entities prefer to lock in a rate, others prefer to have it float. So they swap those risks.

Well, leverage is kind of a separate is from whether or not your entering into a derivative contract or making an investment in the underlying asset. Bu the simple answer is, the more you lever up any investment, the more money you stand to make.

Or lose.

"But of course newspapers and grotty rags like to perpetuate this narrative that derivatives are going to blow up the world."

It's no secret that Warren Buffet is strongly against these derivative products and considers them to be one of the most dangerous financial products in history.

So why is he "perpetuating the narrative" as you dismissively claim?

I want to bring him up because you're hiding behind attacking the newspaper and "rags", so let's instead point to one of the most successful investors in modern history who agrees.

Why is he wrong? Please don't hide behind dismissal and attacks, as a man of Warren Buffets caliber has proven through real world demonstration that he possess a vast understanding of these markets and products.

Something as simple as a swap is a derivative. Swaps are common and no doubt some of Warren Buffet's businesses use FX swaps to avoid being exposed to currency fluctuations.

What would be intensely interesting would be if that derivatives number was broken down by type of derivative. That would be something worth discussing.

Buffett trades derivatives.

Buffet doesn't like what he can't value.

Good, concise illustration from Reddit:

"If Powerball sells 1 million tickets at 2$ each, where the jackpot is worth 200,000,000$, is the size of that market 2M$ or 200 trillion $" [1]

[1] https://www.reddit.com/r/Economics/comments/3xb2xh/all_of_th...

I'll note that the article doesn't have a giant box for the combined value of all insurance policies, maybe because it's obvious that having $100,000 liability policy for your car insurance doesn't sound technical and scary like "derivative."

What blows my mind is that it seems like everyone in the media has taken this obviously incorrect premise and run with it.

The problem with derivatives is that they are functionally very similar or identical to insurance. But, because they are sold under a different name, the capital requirements for selling insurance do not apply. Insurance is less scary for a reason. Higher capital requirements, and less concentration of risk, means that insurance can't blow up the world economy.

This is not what's happening though AFAIK. In the lottery illustration there will be a single payout. In derivatives the risk of default for a certain "class" is treated as IID, and banks can leverage a certain amount on different risk classes (100x is not unusual).

But as 2007 demonstrated the real world risk is not IID. And at 100x leverage a 1% movement can totally wipe out a bank.

Also it's not like the 2007 crisis was fixed in any meaningful way. The FED just bought 1.4 trillion of MBS derivatives and is now sitting on them, actual value still unclear. Probably to be paid by the tax payer through 20 years of austerity.

But this has nothing to do with notional value. The world wouldn't be safer if we rewrote all contracts saying "1% on a notional amount of $100M" with "1M% on a notional amount of $100".

That's not a market at all. That's a $198,000,000 giveaway.

Well, not quite -- if odds of winning are in fact around 1 in 200MM or less, there is no market failure here.

Their point would still stand if they sold 100MM tickets though, so as not to confuse the two issues (ie. the notional amount and the expected value), perhaps it would have been a better example that way.

Only if someone is guaranteed to win the big prize.

In a one-off lottery scheme, the jackpot of a lottery should be less than or equal to the value of all of the contributions.

Did you introduce a central bank into the example some point? Is this a repeated game? Am I missing something? If not, this reddit comment makes no sense whatsoever.

You're assuming someone has to win.

If this was true, Lehman Brothers would still be in operation today. Derivatives, by their very nature, are harder to truly evaluate in terms of value, and the chance of them going horribly wrong has already been demonstrated as being non-zero (see GFC).

Lehman went bust because liquidity in the credit markets dried up - i.e they were unable to roll short-term financing to fund longer-dated balance sheet. The markets thought Lehman had big exposure to the US residential and commercial mortgages, hence they couldn't borrow, hence they went bust.

A shame because they weren't actually bankrupt - PWC has now recovered 100% of liabilities with assets to spare.

How long did the recovery take? PWC recovered those assets now, rather than then. In the mean time there was a huge reflation of the stock and housing markets, powered by the Federal Reserve in the short term and a return of confidence in the medium term.

You can't treat mass psychology as an externality. Austerian policies rely in part on disciplining debtors to restore confidence in market transparency and the rule of law. If thousands of households lost their homes when their cash flow could not meet their debt payments, why shouldn't a bank go bankrupt when its cash flow cannot meet its debt payments? Lehman was just the bank that couldn't find a seat when the music stopped, and that's fair.

Yes, but that has nothing to do with the notional value of derivatives or derivatives "blowing up." They had a liquidity crisis because of fear; people stopped trusting their commercial paper and were skeptical of lending to them with repos backed by difficult-to-value mortgage-backed securities.

"The markets thought Lehman had big exposure to the US residential and commercial mortgages, hence they couldn't borrow, hence they went bust."

This is the sort of thing I mean when I say derivatives are hard to evaluate correctly. So yeah, Lehmans totally went bust because of their large amounts of trading in CDS.

This. Makes me question the rest of the graphic's accuracy.

This is exactly right. People are confusing risk with notional. There's no risk adjustment anywhere, but clearly the values need to be adjusted for how much they might change.

The cash will be (by construction) the same tomorrow, but the shares could easily vary by a few percent.

In a more ELI5 sense, would it be true to say that derivatives are not "real money" in the same sense that company valuations are not?

Company valuations are projections created through some combination of extrapolation of growth or value and (often in the earlier stages) just agreeing on a number to make the percentages and such work out correctly. They are expressed as a dollar amount but those are not real dollars.

Normally you're right that no net exposure implies no risk. However, you've completely forgotten the financial crisis? Counterparty risk can completely destroy the idea of "all positions are offset". If I buy a derivative and then re-sell it for a profit, that's fine until the people I bought it from suddenly go under, and now my position is not protected anymore!

Comparisons of gross national to other measures of value or risk aside, isn't the point that the risks inherent in this magnitude of derivatives trading are simply not understood? "Blowing up the world" might be hyperbole. On the other hand, these markets are mostly unregulated. Does anyone really understand the systemic risk posed by this segment of the market?

I agree with this, but let's not forget about the counterparty risk making the interconnected financial system so fragile. If GS offsets its [long position with JP] with a [short position with Citi], and Citi goes under, GS is screwed.

yeah but most interbank dealing is collaterised now so little counterparty credit risk left (altho this does arguably just shift counterparty risk to funding risk, but that's another story...)

Any book recommendations for someone without a background in finance who wants to understand more about derivatives and related concepts?

Veale is one of the widely recommended authors: http://www.amazon.com/Stuart-R.-Veale/e/B001IOBOR0

specifically this book: http://www.amazon.com/s/ref=nb_sb_noss?url=search-alias%3Dst...

Well, to be fair, derivatives almost did blow up the world in 2008. Definitely overestimated in terms of impact, but really something we should have a better handle on.

Except that derivatives were not the cause of the financial crisis. Over-enthusiastic mortgage lending to poor borrowers caused the financial crisis.

Over-enthusiastic mortgage lending to poor borrowers without risk, because the risk was wisely managed and diversified through perfectly designed, financial, innovative products that covered all the bases (for a nice commission, of course).

The mortgages everything was based on were junk, because they were guaranteed by the GSEs that couldn't handle the rate and size of the defaults on junk mortgages.

Everything based on junk mortgages were junk.

Yes but those products (Mortgage backed Securities) are cash products, like a bond or equity, NOT a derivative.

Depends on the definition. The wikipedia editors think it is:

https://en.wikipedia.org/wiki/Derivative_%28finance%29#Mortg...

Derivatives did exacerbate the unwinding of the mortgage mess, because the chain of title was so convoluted. I've read reports of people who were given homes in south Florida because the owners of mortgage debt were unable to finish foreclosure proceedings before the statute of limitations ran out. That is an extreme case, but the milder case of banks being landlords when that's not what they're good at, plus borrowers being stuck in foreclosure limbo longer than should be necessary, introduces unnecessary friction in the economy that wastes time and money.

This could be the self-regulating limit for derivatives - the legal time required to unwind them in the case of mass redemptions, bounded by the business cycle length and statute of limitations for a locality.

Weren't the banks selling each other credit default swaps they couldn't cover, such that the moment a single bank went under it triggered the collapse of the entire banking system? Blaming the mortgage lending is like saying, "I didn't blow up the neighborhood with my unsupervised warehouse full of explosives, it was those damn kids playing with matches!"

Some economists call this view "The Big Lie". See Krugman's column today: https://www.nytimes.com/2015/12/18/opinion/the-big-short-hou...

That is one part of it. Those mortgages could be blended into derivatives, which could then be run up in a speculative bubble. Pretending there was a single cause for the financial crisis is a little disingenuous on both our parts.

what does "blended into derivatives" mean? Yes mortgages were sold on a packaged up into MBS (Mortgage back securities), but these are cash products, like a bond or equity, NOT a derivative.

Ahh. I stand corrected. Looking at your later post (which for some reason I can't respond to directly): so the derivatives are just what amplified the MBS crash with leverage -- allowed organizations to put more money into toxic MBS assets than they would normally be able to with cash? All the splitting and recombining of MBS does that not act as a vehicle for laundering or coverup? I mean it wasn't technically laundering because it was completely legal to offer mortgages to people who couldn't afford them. But didn't the MBS allow banks to resell those poor assets which normally wouldn't have been sought after as investments? And the investments were leveraged through derivatives so that crashing of the poor quality assets were amplified? So it seems that there was a complex interaction of unregulated financial vehicles that caused the crisis and not simply "bad mortgages" or simply "derivatives" or simply "MBS". Any one of those alone would not have caused a huge bubble and crash, right? I do not have a background in finance, but I think it is important for a layperson to understand what happened.

There were MBSs built by splitting bundles of other MBSs into various tranches. Those certainly seem to be derivatives.

What would they be called if they aren't derivatives?

> What would they be called if they aren't derivatives?

Securitised bonds, which is a cash product.

The distinction between a cash & derivative product is really simple - one is leveraged the other isn't. E.g if I buy 100m of MBS I actually have to cough up 100m in cash - I am buying both the interest and principal repayments of the underlying mortgages.

On the other hand if I enter into an interest rate swap with 100m notional I don't pay anything at all (assuming traded at spot market levels, and ignoring initial margin). The risk, or sensitivity, to interest rates is the same on both, but am only paying out cash for the securitised bond.

Hence the bond, or MBS is "cash" and the swap is "derivative".

Portfolios. There are synthetic MBSs, too. Those are derivatives. They were also relatively uncommon.

I was thinking the exact same thing and came here to point this out.