A: A pairs trade involves buying the 'low' side of a pair of stocks and short-selling the 'high' side of the pair. As the gap between the two securities closes, you stand to make a gain, even if both are falling. Spread bets offer great flexibility to be used in this way; not only do they cover thousands of shares and entire sectors of the market but they are also traded on margin which means that you can leverage your gains. You could even buy, say the retail sector while short-selling the mining sector.
Taking a view on a single stock can be difficult - not everyone is comfortable deciding whether Tesco is fundamentally underprices, or overpriced, for example. However, with a pairs trade, you don't need to worry about that or even whether an entire sector will rise or fall. Instead, you take a view on whether one stock is cheap or expensive in comparison to another. So you could do some past research into the price relationship between stocks that are heavily exposed to the price of oil, such as BP and Royal Dutch Shell. Once the two prices move out of line, you buy one stock and sell the other. You may, for example, not know which way the oil sector is heading, but be fairly sure that BP is undervalued relative to Shell. So you could buy BP at say £7 per point when the price is at 400p and sell Shell (RDS) at £2 per point when the price is 1400p. The total notional value of each trade is the same (£2,800), but if BP now performs better than RDS, you make a profit. If you close out your positions when BP's share price is, say, 480p, and RDS's is 1,625p, you would make £560 on your BP trade - (480 - 400) x £7 - and lose £450 on RDS (1,625 - 1,400) x £2. That's a total profit of £110. This ignores spreads on both legs of the trade, but these should be fairly small.
|
Share |
Stake |
Total |
Share |
Profit/Loss |
|
BP - buy |
400 |
£7 |
£2,800 |
480 |
+£560 |
RDS - sell |
1,400 |
£2 |
£2,800 |
1,625 |
-£450 |
|
Total Profit/Loss |
+£110 |
||||
You can also use pairs trades if you think that one index is overvalued compared with another. Let's look at the FTSE 100 and Wall Street, for example, when on 7 October 2008 they were trading at around 4,750 and 9,500 respectively. You could have sold £1 per point of Wall Street at 9,500 and bought £2 per point of the FTSE 100 at 4,750, once again giving you the same pound value exposure (£9,500) on both sides of the trade. When at their most recent lows - at the beginning of March this year - the FTSE 100 was trading around 3,500 and Wall Street was around 6,500. Closing out both trades at that point, your profit on the Wall Street trade would have been £3,000, whilst the loss on your FTSE trade would have been £2,500, making an overall profit of £500 (again, before spreads). The table at the bottom sums this up. [Courtesy of Capital Spreads]
This works by simply buying the 'low' side of the pair and short-selling the 'high' side of the pair. As the gap between them closes, you make a profit, even if both are falling. As always with spread betting, the safest approach is to enter the trade once the move you're hoping to ride actually begins. Just because one price has got way out of line with another one isn't a reason in itself to trade. After all, the relationship could get even further out of kilter before it snaps back.
|
Share |
Stake |
Total |
Share |
Profit/Loss |
|
FTSE - buy |
4,750 |
£2 |
£9,500 |
3,500 |
-£2,500 |
Wall St - sell |
9,500 |
£1 |
£9,500 |
6,500 |
+£3,000 |
|
Total Profit/Loss |
+£500 |
||||
A: That's called pairing and is the only way to trade forex...
e.g. dollar V pound is called cable...you trade dollar and euro... etc... i.e. if you are long on the pound you are short on the dollar...
For instance at present the FTSE 250 (trading at 15 times next year's projected earnings) looks the more expensive index when compared to the FTSE 100 index (which has remained stuck at 12 times earnings).
So if the long-run relationship between the two indices is to be re-established the FTSE 100 will have to outperform the FTSE 250 in the future. Thereby you can place a spread bet on the blue-chip index rising and at the same time place an opposite spread bet that the FTSE 250 index will decline. If you are right about the relative performance, the profit from one bet will be bigger than the loss from the other and you will book a net gain.
What is great about this trade is that irrelevant of whether the market as a whole goes up or down you still stand to gain from the relative performance.
Of course you still need to make sure that your exposure is the same on both bets, so if you bet £25 a point on the FTSE 100 at 6,795 you need to offset it with a bet of only £14 a point on the FTSE 250 at 12,126. Imagine both indices moving by 10pc either way and you'll see how this works.
As the spread bets are balanced it doesn't matter whether the FTSE moves up or down. An equal percentage rise or fall in both the FTSE 100 and the FTSE 250 leaves you with a profit on one trade and a loss on the other that cancel each other out. It is what's called a 'market neutral' strategy. Your net gain or loss comes purely from the relative performance of the two indices.
Market neutral is doesn't mean that this technique is risk free. Even though it doesn't affect you whether the market rises or falls you can still lose money if you are wrong and the FTSE 250 continues to outperform the FTSE 100, in either a falling or rising market.
A: What you're doing is essentially arbitrage. I've executed a similar trade not so long ago involving taking a long position on the Dow and a short on the S&P aiming to exploit the difference between the two. The same could have been done with the FTSE and the Dow over the last month or so, as the Dow was lagging behind by some way.
Exploiting these differences is called arbitrage. A lot of people also like taking advantage of the difference between the futures contract of, say, the Dow Jones, while buying stock for the same sized position in Dow Jones equity. It involves double stake thinking, but obviously it has to be exact for it to work....or near enough at least. If you were £1 per point on the Dow Jones, you'd divide that position size by the FTSE, which would give you the required £ per point there.
It can become very complex indeed, which is why people program computers to do this kind of trading for them but, you can just as easily do it with two futures contracts (the security followed by spread betting companies when looking at the Dow or S&P). The computerised trades make up roughly 75% of volume on the New York Stock Exchange if I remember correctly. These programs buy and sell huge volumes over very short periods of time and basically scalp tiny profits whenever they get the chance. There are people making a good living from the strategy you use. It's definitely low risk and it definitely works, so keep it up if you're interested.
A: Using the closing prices of PRTY (210p) and 888 (158p) the straight pairs trade would be to buy PRTY @ 210 and sell 888 @ 158 (or where ever they are trading during normal market hours).
If you traded one for one then you might buy £10 pp of PRTY and sell £10 pp of 888. For every one pence move of either equity you would win (or lose) £10.
PRTY price moves to 231p and 888 moves to 174p.
Both have moved by 10%.
You have made 21p x £10 = £210 on PRTY.
You have lost 16p x £10 = £160 on 888.
Your net win is therefore £50.
It is worth noting that your weightings per equity may not necessarily be equal which could potentially skew your trade.
You may want to consider trading by notional amount eg you could trade both equities on a notional £2000 basis. The stakes you would trade would then be approximately:
2000/210 = £9.5 buy PRTY.
2000/158 = £12.6 sell 888.
If the price of both companies went up by 10% your net p/l would remain relatively flat. However if you anticipate PRTY outperforming 888 and the reverse happened at least you would be covered (relatively).
A: To address the first part of your question, yes this is a sterling neutral transaction.
Secondly, see the table below to explain why your account will not necessarily be the same.
| Going long and short by equal amounts on two highly related stocks | ||||
| Day | Long | Short | Profit/loss per share | |
| Day 1 | $100 | $100 | 0 | |
| Day 2 | $102 | $101 | +1 | |
| Day 3 | $105 | $102 | +3 | |
| Day 4 | $98 | $92 | +6 | |
As you can see on Day 1 the spread bet account was neutral but as the shares move on Day 2-4 the spread better will make a profit or a loss as the correlated stocks usually do not move in the same ratio.
In response to the third part of your question, there are five different scenarios with a pairs trade.
(1) Both stocks could go up
(2) Both stocks go down
(3) Long stock goes up, short stock goes down
(4) Long stock goes down, short stock goes up
(5) Both stocks stay the same
The scenarios you want are (1), (2) and (3). You do not want scenario (4) or (5).
A: Ok, in a nutshell with a spreadbet you are trading with a sterling stake even if the stock is denominated in a different currency therefore $3.91 represents 391 points and Tullett Prebon is currently 2.61 therefore 261 points. With this logic Tullett Prebons' price is roughly 66% of GFI Group therefore your stake in GFI Group would need to be roughly 66% of your stake in Tullett Prebon.
Further Explanation: This is trading purely on points and quite a number of serial spread bets are placed in this way. Here, you are basically taking a hedging view that one company will outperform the other (trading based off price correlation) and lowering your risk and BUYING the better while SELLING the lesser performer. When trading a correlation serial bet you are making more of an assumption that one of the stocks is overpriced/underpriced against the other.
A: To find the correlation the easiest way to do it is to have a Bloomberg or Reuters terminal which you can run quick data analysis but these cost an arm and a leg per month. The other way is to download historical prices from Google Finance and run the formula in Microsoft Excel: =CORREL(,) from the two price sets. Here is Tullet's data from Google Finance: http://www.google.com/finance?q=LON%3ATLPR and you can see just above the graph on the top right it has a historical prices link. Click on that and download the data for however long you want to back date it to. Do the same for GFI run the prices into the formula. If you need help with how to use the formula read the Excel help manual.
Below is a graphical illustration showing the correlation between the two stocks. It isn't of the greatest quality but we can interpret the figure and as of today the correlation of the two stocks reads 0.5651 meaning for every $0.5651 GFI moves Tullet Prebon will move £1. From this we can gather that the two stocks are moderately correlated and we would trade them at a ratio as follows:
0.5651 GFI to 1.0000 Tullets or...
1.0000 GFI to 1.7695 Tullets (1/0.5651 = 1.7695) it doesn't matter which way you do it, it is still the same ratio.
 |
The 66% percentage mentioned in the earlier question in a way is trading correlation though on more of a rough scale. People do trade this method but mainly on indices because there is more of a historical basis for the spread. For example over the past 20 years the DAX index historically trades roughly 200 points above the FTSE index. Therefore you could easily exploit this if the spread goes to 400.
There is another way of doing it based off nominal value assuming that the spread you are trying to do is a BUY/BUY (or SELL/SELL) spread rather than a BUY/SELL (or SELL/BUY) spread which means you are BUYING (or SELLING) both securities at different ratios rather than taking a hedging view that one will outperform the other (as in the above example). With a nominal value serial bet you are taking advantage of a mispricing of fundamentals.
Let's take an example of a nominal value serial bet. We first need to know each price of the stock in their base currencies.
GFI closed trading at $3.91.
NB: This trade takes the assumption the ratio would never change but we know this is untrue due to currency fluctuations as well as price asymmetry.
So technically you are making a bet on two variables:I will explain the currency fluctuation using an example of buying physical shares. When buying physical shares in a regular market transaction you pay the nominal value i.e. number of shares x price per share.
Take our two companies for example. Let us say we want to buy 100 shares of each.
100 x $3.91 = $391 nominal value of GFI
100 x £2.61 = £261 nominal value of Tullets
Seeing as though these two stocks trade in different currencies the only way to find a comparable nominal value is to convert it to a base currency. So we convert GFI to GBP as we want to trade GFI in GBP.
Below is also a Bloomberg screenshot, which calculates these ratios as explained earlier. Unfortunately the screenshot didn't come out too clear but I will quickly run down what it shows.
 |
In the top graph you can see the prices of both companies GFI in white and Tullet in orange. I've converted GFI to base currency of GBP on the top right of the page (as we want to trade it in GBP).
The last trade of the day for GFI converted into GBP was £2.687535 ($3.91). Tullet's traded at the close of £2.61.
Turning our attention to the graph at the bottom this shows a ratio of the two stocks having been converted in the same currency. In the little yellow box in the top left of this graph it says 'Ratio: 1.0297' which can be calculated by dividing the new GFI price £2.687535/£2.61 Tullet's = 1.0297 (or 1.03 rounded up).
A correlation bet is made on how you think two stocks will move in conjunction to each other. Stocks with a correlation of -1 (the lowest possible correlation) will move inversely against each other 100% of the time where as stocks with a correlation of 1 (the highest possible correlation) will move exactly in line with each other 100% of the time. Anything in between is up to your interpretation. The nominal bet one is only common for those who are more experienced investors who try to take advantage of things like dividend arbitrage and takeover arbitrage, which is getting to the stage of quantitative analysis and above my head.
The practical use of buying two securities at different ratios is as such. If you know both stocks are highly correlated (usually >0.5) then if you are more risk averse you would put your lions share in the stock that makes bigger moves i.e. the more volatile one. If you aren't so risk averse, put the greater percentage in the less volatile stock.
To lower your risk there is a good trading technique called hedging, which is to buy one and sell the other (described in the earlier questions above). This method won't reap you as big a profit (or loss) as buying (or selling) both but you know that if both are correlated and move the same way one will profit and one will lose every time. Therefore there are only two ways you can lose 1) is if the two stocks become less correlated to the extent that they become uncorrelated or 2) the more volatile stock moves less than the lower volatile stock.
A: Yes, of course you can do this using a pairs trade... However, if considering this route you should always try to balance the two positions into equal bets. For instance going long £10 on the FTSE and shorting the DAX at £10 a point is not an equal and opposite trade.
What you can do is to take the current values of the FTSE 100 (say 5400) and Dax (6000) and divide one by the other - in this case the ratio would be 0.9 to 1. So you would then trade £0.9 points on the Dax for every £1 on the FTSE 100.
Note that you might need to round the numbers as some providers might not take 0.9 a point.
0.9:1 is equal to 45:50.
Swings on the Dax can be more extreme than the FTSE 100 index and thereby the potential profits/losses are greater. The Dax is made up of 30 German companies as opposed to the FTSE 100 meaning that earnings by individual companies or sector have a bigger impact on the Dax than the FTSE. Not only that but the Dax moves rapidly compared to the FTSE 100 when movement in the Dow Jones/ S&P futures happen so it will not come as a surprise that this is a very common trade amongst speculators. I occasionally trade both the FTSE 100 ad Dax and often find that the FTSE is sometimes lagging its European counterparts which gives you that little bit of extra confidence when placing orders either long/short.
The content of this site is copyright 2016 Financial Spread Betting Ltd. Please contact us if you wish to reproduce any of it.
