One of the most misunderstood concepts in all of finance is that of “arbitrage”. It looks very much like speculation from afar but it isn't, it's very much the opposite. There are also a lot of people who describe what they do as arbitrage and they're damned liars: they're speculators. Ivan Boesky used to claim to be an …
Here's one for the common man
Tesco is selling gold jewelery at less than the bullion value (London Fix) of the gold in it.
They don't actually advertise this, and probably don't even realise it, but I bought a 9 carat gold bracelet from them for £27, which when weighed came to 2.1 grams. That contains 0.7875 grams of pure gold, which at £1145 a troy ounce is worth £28.99. Plus you get a nice presentation box.
So I bought four more.
They are still available:
Buy two and use the voucher code TDX-KLH4 to get an additional £10 off. Also collect from store to avoid the £5 shipping charge.
And just how exactly are you going to realise this gain? I suspect you're going to feel a bit of a mug when you pop into Cash Converters and they explain that the rates they offer don't *quite* match the international bullion markets...
How do you propose to extract bullion-quality gold from these bracelets without paying more than the price discrepancy? (Note: the cost you'll have to pay includes the assaying necessary to get it marked as really bullion-quality.)
Nice if you can sell it at London Fix, the best price I can find is £27.93 for 2.1g. Still seems a fair deal though with the discount, thanks for the tip!
"Nice if you can sell it at London Fix" et all
Fair enough that it isn't truly arbitrage if you don't realise the gain instantly with a corresponding deal. However, my method of releasing the value involves an intrinsically long-winded transaction. If you don't actually know five girls, this may not work out for you (or just buy one or two).
Maybe you're not talking to the right girls ;-)
You're also forgetting
Clubcard points! I reckon you've found a way to bring down Tesco, nay, the world economy!
Yes, I get clubcard points, and I link through to Tesco Direct via TopCashback for about 3% more discount, plus I pay with My Nectar credit card so I get Nectar points too.
But that would have made the first post a bit lengthy.
re: Clubcard points
I think you missed my sarcasm. The price you've quoted for bullion gold takes into account what it costs to turn regular gold into certified bullion gold along with a multitude of other factors.
In short, 9ct gold does not have a value 37.5% of bullion gold.
re: re: Clubcard point
Yeah, sarcasm is hard to get across- I took your post as a rather lighter form of humour.
As a jeweler, I actually used to pay MORE for carat gold than the bullion value of the contained gold, when buying it as a raw material. And the people who bought it from me subsequently (in jewelery form) paid a lot more. Shop markup on my prices was normally 100%. Purchase tax in those days was 35%.
Buying gold as jewelery is a traditional form of savings in some cultures, despite the swingeing markup. This Tesco deal lets you buy at the actual gold value (actually a little less). The article should be worth more than the metal content when you go to sell, just don't insist on selling at scrap value- sell it as a gold bracelet.
There are many ways to invest in gold other than laying up .999 bars. Some people prefer coins, which trade at a premium over their gold content, even though they are often not pure gold. In the coin market, many people prefer to trade silver coins rather than bullion, because silver bars attract VAT on the whole value, whereas trades in coins attract VAT only on the trader's markup (there is no VAT on gold bullion or coins, but there is on jewelery of course).
transaction costs and Tobin tax
The reason arbitrage as described in the article works is that the arbitrageur pays almost nothing to buy and sell in 2 markets simultaneously. This never applies to real stuff, like gold bracelets for reasons stated in other responses. It only works on electronically tradable virtualised securities, until governments decide collectively to introduce the Tobin tax on such trades. Doing that would force speculators to take a long enough view of something to help the change in value to be more real as opposed to short-term and psychological, and it would limit arbitrage to situations where the price difference is big enough to pay the Tobin taxes on the simultaneous deals.
Boesky did arb...
Arbitrage is normally done at the same time between different markets, but can also be done between different times in the same market. If I know something the market does not, I can reliably predict the future price of something, or at least the direction of change. So I can arb between the price now and the price in the future. There is a tiny risk, of course, in the sense that maybe the market knows more than I think about the item, but there is also risk in cross-market arb, in that the price might move sooner than I thought, and in the wrong direction.
What Boesky did *was* arbitrage in this sense, but he used inside information to gain that knowledge advantage, and that was ultimately his undoing.
"So I can arb between the price now and the price in the future."
No, you can't. When you do that you are taking a position. The arb implies that you have no overall position.
Depending on how exactly you are doing it you may speculate on outrights (the highest possible risk) or on time-spread (usually considered lesser risk but all depends on which particular markets you are talking about).
If you are talking about commodities you can use the strategy you mention to hedge your equal but opposite position in the physical goods but even then it is not an "arb". Normally such things are known as "contango plays" - when the forward curve is going up sufficiently steeply, you can buy your physical goods now, place them in storage and sell futures for the same quantity, say 3 months forward. If the contango is large enough to cover your financing and storage costs you will make money.
But even in that case there are still risks involved, higher or lower depending on whether the futures you used are deliverable or financially settled, whether your goods are of exactly the same quality, same location as the futures delivery basis etc etc.
Or you could call it what it is...
Which is speculating...
Hes speculating the price would rise/fall - You don't *know* as it hasn't yet.
"If I know something the market does not"
that's called "insider trading", not "arbitrage"
I wouldn't really call BT traded in London versus NY arbitrage as you are, say, short an NY stock and long a London one. You may have locked in this price differential but at some point you need to close out unless the NY share is fungible with a London one. Otherwise you are carrying fx risk until you close out.
A better definition of arbitrage is the future on an index being a function of the index and borrowing/discount rates. Even then risk-free is only theoretical as you have execution risk - the ability to get both sides of the trade on at the prices and in the sizes you need. Due to this prices vary by an amount reflective of this and the ability to borrow cheaply (close to the "risk free" rate)
A lot of what is listed as arb these days is just big basis risk due to playing around with proxies (gold mines/shares vs bullion) as true arb opportunities are thin on the ground.
Very interesting. Thanks!
Arbitrage isn't entirely risk free
Arbitrage also tends to boil down to a big game of "who's fastest?". Say I've noticed that I can buy BP shares in New York for the $38.50 and sell them in London for (the equivalent of) $38.52. I send off my orders to New York and London at the same time. Unfortunately, whilst I'm doing that, someone else sells to the person in London who's offering $38.52. Now that they've done that, the best price available in London is now $38.49. My order in New York gets executed fine, but I'm now left with a bunch of BP shares that I can't sell for the same price I bought them for, never mind for a profit.
There are ways and means of trying to deal with this and similar risks, but none of them are perfect.
This kind of risk is fairly small, but it's nevertheless not true to say that arbitrage is risk-free.
"This kind of risk is fairly small, but it's nevertheless not true to say that arbitrage is risk-free."
This is an execution risk and you run it *before* your arbitrage is "locked". After that you should be risk-free or you cannot call it "arbitrage".
But note that buying the same product in one market and selling it on another is not risk-free unless you can move your product from one market to the other to liquidate your trade. If that is not possible then you are still exposed because you will have to liquidate your positions in each market separately and the spread between them may move against you - so it's not a true arbitrage.
I once had the pleasure (ahem) of working on training courses for the New York Stock Exchange an rather wish some of the writing on those had been as clear and succinct!
Yeah, Finance is obsfucated to try to make it sound like a mystical art.
I believe obfuscation is buried somewhere in the definition of Profession. It may not be explicit so possibly it would be best to ask a lexicographer.
Paris, who isn't obfuscated in the slightest.
Obfuscation, acronyms and other ways of sounding smart
Paul Wilmott does a good job of debunking the bullshit and I would certainly recommend his "introduces quantitative finance" series
"obsfucated to try to make it sound like a mystical art"
I could say the same of medicine, law, or dare I say it software engineering...
I confess to being a bit of a nitwit when it comes to economics, and I can't get my head round the following:
With all these hot-shot traders moving stuff around at breakneck speed, and making a handsome living at it, surely for every top dog killing there's a numpty who's made a loss? If so, where are all the fallen?
I presume those are the ones who didn't trade fast enough and are left with billions of pounds worth of stock, brought at higher prices than its now selling for.
It helps to not consider that the world economy as a balanced equation. If we froze the world now, the banks, the assets, the debts, etc, and settled everything, it would not return to zero, or even the "value" of the assets.
Got a pension fund - that'll be you. Any savings - same deal. The real distinction isn't between retail and investment banks but between banks that offer real products and services into the real economy and the arbitrageurs, hedge funds and so forth who do nothing for anyone but them selves. That explains the rewards that are so unconnected to the real world. The competition isn't to provide those goods and services, it's to find the sharpest dealers, the smartest quants, the crookedest tax lawyers and so forth. And all the while anyone who operates honestly in the real economy is being robbed. Bring on the Robin Hood tax, the Tobin tax, whatever you like to call it. A very small percentage would still throw enough sand in the works to stop these shenanigans.
Not a zero-sum game
Not only the economy as a whole is not a ZSG but also each market on its own isn't either.
Arbitrage is valid work. Look it up if you don't believe.
"Die im Schatten..."
> where are all the fallen?
No longer taking part in the general discourse, becoming - at best - part of a statistic no one is interested in.
That's why the current turbo-capitalism looks so good: Winners only.
" There are some who are in darkness
And the others are in light
And you see the ones in brightness
Those in darkness drop from sight "
If you don't know who the mark is...
then It is you.
Said of poker games. Analogy of markets, where the little fish with IRA and 401K (or the UK equivalents) assets are called upon to cough up money for the professionals. The little fish think they swimming with the sharks and the little fish are getting a little food, but there comes a day when it's shark feeding time.
The depiction of a bookmaker is not incidental. There exists leveraged ETF's for both sides of an index, long and short. The firm makes money up or down, uses the losers money to pay the winners, pockets administrative fees and the fractional losses in daily volatility. Check out FAS and FAZ on NY market. Bookmaking, but it is "financial" and not sports betting, so it is legal in NY.
There was an item on the R5 "wake up to money" program last week (n.b. I listen to the podcast - I'm not awake at that time of day!) about a company that was laying a new dedicated cable across the atlantic for a finance company ... by having their own direct transatlantic connection they reckoned that they'd get data between NY and London 10ms faster than at present and that for every ms gained they thought they'd get a return of something like $100million. So speed definitely is important!
I work for a very large telco providing very fat pipes for most of the worlds major trading houses and the thing that upsets them most if there is a fibre cut is not the 50ms time it takes to switch to the protection path but the 5 - 20 ms increase in latency!
These brokerage firms are often some of the fastest ways we find out there has even been a fibre cut somewhere in the network because they'll raise a trouble ticket because of a latency increase before the network monitoring center ever pick it up (it being just 1 more blinking alarm in hundreds of thousands of other alarms!)
I'm told that those trading in New York can connect either over the internet or by having their automated trading systems in a rack at the stock exchange. The latter gives an advantage and therefore you pay big bucks for a rack. I'm also told that some European exchanges forcibly add random delays to transactions in order to avoid this problem.
Perhaps some IT specialist needs to teach the bankers about race conditions.
Have they looked at neutrino beams?
Very nice article. It explains something interesting in a very understandable fashion. Well done!
Great article! But I have spotted a flaw in your analysis -
You describe the process of arbitrage in a way that is mathematical and involves rational calculations and assessments, implying that these positions are primarily staffed by responsible, intelligent people.
The reality to most people who know traders is that they are cocky, arrogant, ham-fisted cretins, too high on coke and strippers to remove their heads from their rectums far enough to even look at anything approaching a rational calculation.
But then that is the way it is supposed to work. Arbitrage should be a matter for the quants, risk taking for the traders.
That it often isn't is one of the problems that finance has managed to get itself into.
Ah, something I can upvote.
But why don't all Register articles have a "rate this article" feature? Some of the "Reg Hardware" and "Channel Register" etc. articles do, but the front-page articles don't.
Anyway, thanks for an excellent article, Tim Worstall.
Is "Tim Worstal" "Tim Worstall"?
All the traders I've ever met have been cocky and arrogant, certainly, and who doesn't appreciate coke and strippers? (well, maybe those who prefer hands-on entertainment...) but I certainly wouldn't call them ham-fisted or transanocephalous.
If they've accomplished anything at all, it is because underneath it all they tend to be at least moderately capable. They wouldn't ever get anywhere if they weren't, and those hookers don't come on credit you know...
Not long ago you could buy oil on the spot market at one price and sell it on the futures market at a higher price. The deals went through at the same time, but then you had to hang on to the oil somehow until the date for the future delivery. They had tanker ships steaming in circles for a month.
Oil AMD commodities in general are unique in that the future price can be less than the current. The reason i believe being the risk that you cannot deliver the goods - other futures being cash settled. This risk of non-delivery comes and goes giving rise to contango and backwardation situations.
"surely for every top dog killing there's a numpty who's made a loss? If so, where are all the fallen?"
Probably our pension funds!
Nice article but many facts simply wrong
"A bookie will always have a balanced book: whichever horse comes in, the house wins (same with casinos). "
Having worked for a major FTSE bookmaker for many years your bookmaker (and casino) analogy is quite simply totally wrong. Bookmakers in reality about 99.99% of the time never have a balanced book on an event, it's where the expression 'the bookies took a beating on that one' comes from.
To be correct you should have said "Despite bookies almost never having a balanced book on an individual race they price the available odds to ensure that they have a long run 'edge' built into all their books to ensure a long run profit margin, however significant day to day volatility can see the daily margin swing wildly from negative to positive numbers"
For example lets say I price my horse race to have an overround of 10%, in English that means that I've basically moved all the odds down such that I have a 10% mathematical advantage(the overround) in the book. If (AND IT'S AN ENORMOUS IF) £100 is distributed across all the horses in the race as per the actual underlying odds I'll have my balanced book andf make £10 whatever the outcome.
However life's not like that in reality, about 60% of the money ends up on the favourite and about 40% on the rest of the field. So if the favourite comes in what tends to happen is that the bookie actually truly, really loses money on that one race and the books 'actual' margin goes negative usually to something like -20%.
Howveer lets say the favourite goes down at the first fence, the bookie then makes out like a bandit and gets a margin of +say 40%, on that race.
Over many races the ups and downs smooth out and most bookies tend to target a horse racing margin of around 15% over the long term.
It's a much misunderstood area and requires bookies to be very careful about their own risk hedging, there's always a potential daily loss limit that will be tolerated before the bookie himself starts hedging off (usually back onto Betfair or between each other).
It's worth noting that the Tote system works differently as it uses the weight of money on each horse itself to calculate the underlying odds and then takes a rake from the pot of money as it's fee.
The Tote has no risk from the race itself as it's fee is independent of any combination of outcomes (ignoring cancellations, Acts of God etc.), but it's not balancing it's book as technically it's not operating a 'book'
Arbitrage is also not totally risk free as you describe, there's a risk that your buy and sell orders don't execute properly, liquidity is not sufficient to cover one side of the trade or the other, counter party risk etc. etc..
Arbitrage can be very low risk but it's not NO RISK.
Hope that helps some people.
The hobby horse won THAT race
To be fair, if Tim had written up his bookie analogy to that level of detail, I'd not have made it to the end of what was an excellent article.
Good point - the bookie analogy is a simplification and omits the crucial factor that he cannot change his odds retrospectively. let's suppose that he has taken some bets at 10-to-1 on a horse, and there is a rush of money onto that horse just before the start. He cannot whack the odds down to evens to cover his previous exposure, well not unless all his bookie chums do the same - but then large adjustments generate all manner of instabilities. So, he has to ride-out the ups and downs, to take a real position on the race, and in reality he is a speculator, albeit a reasonably well-hedged one.
Excellent Article notwithstanding.
Probably true but the betting market offers some really intriguing insights into how much of the economic theory used to build trading models is probably totally wrong.
For example: Everyone knows that potential reward rises in line with risk right? It's a key component of efficient market theory and is used practically in how credit card APR's are set, mortgage rates etc. etc.
Well in the betting markets it's completely the inverse for the market participant (read trader)
I'll give you an example: Say there are two blokes who know absolutely nothing about horses. Each starts with £1m and bets on every single race over the year such that they've wagered the full £1m over lets say for argument 100,000 races so volatility will be significantly smoothed out.
Bloke 1 only ever backed the favourite or joint favourite
Bloke 2 only ever backed the longest priced horse in each race.
At the end of the year Bloke 1 will have left of his £1m about £950k
At the end of the year Bloke 2 will have left of his £1m about £200k or even less.
Why is this? Both will have less than they started with due to the bookie's overround but why the difference?
Basically bookies price the favourite at a lower expected margin than long shots (usually only 3-5% off the underlying mathematical odds), so your 1/1 favourite is probably actually about that, whereas that 100/1 long shot is actually realistically about 200/1 or longer.
The bookie doesn't price it longer as you don't attract any more money as the elasticity is too low so what's the point.
You could say betting markets are an anomally and financial markets are different but I very much suspect not.