An amazing filter. - page 6
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
The same can be said for commodity and equity spreads.
For example, the U-contract spread of US securities ZBU1 - ZNU1 = 1:1 (30-year bonds - 10-year bonds - available in mt4 at the moment).
We are interested (I repeat) in areas where the probability of price movements in one direction is 60 per cent or more several times in a row.
We see such an area in the near term - in August. See the figure above.
We are not going to look at the first week - price is falling down at the beginning (red area). But starting from the second week and till the end of August - the spread is likely to grow quite high - 64-73-73% weekly!
Let's see how this expected August "rally" of the bond spread ends up!
I anticipate buying the spread around August 7-8:
BUY ZBU1 - SELL ZNU1 = 1^1
Not a bad idea, provided you diversify your investments across several instruments at the same time. The question is, how credible are the statistics from this company? And how affordable is this case?
Another point. With the current trend of rising food and commodity prices (due to crop failures and natural disasters), statistics can be a let down.
Assuming buy spread around August 7-8:
BUY ZBU1 - SELL ZNU1 = 1^1
No, Mathemat, - for the last month and a half or two months I did not seem to mention anywhere at all the entries on ZB, ZN bonds and their spreads.
Because by seasonality - the last time an entry on the spreads of these bonds took place around May 18 until the end of May. I offered that May entry in Leprecon. Since then, just yesterday I wrote about that spread again here.
As for the current risk of buying the ZB-ZN spread, it will already be known in early August whether the US Congress will approve an increase in the government debt ceiling. Hopefully it does and bond prices lose their current (totally unnecessary to us) "nervous" volatility, enter a calm channel and then it will be "concretely" clear - is it worth it to enter this spread.
Here is an additional graph of the multi-year seasonal ZB-ZN spread lines:
Not a bad idea, provided you diversify your investments across several instruments at the same time. The question is, how credible are the statistics from this company? And how affordable is this case?
Another point. With the current trend of rising food and commodity prices (due to crop failures and natural disasters), statistics can be a let down.
I can't definitively say whether it's reliable or not - see personal info.
----------------------------------------
I don't think food prices are going up much right now. On the contrary!
For example, since mid-July almost all grains (wheat-soya-corn....) have been moving down! Cotton tends to fall.
Coffee is collapsing sharply now in spite of its seasonality (I even caught a moose on coffee KCU1, - unexpected and hurtful - see pic...)
Horned cattle and steers (GF, LC) are also falling, and falling so much that the price is almost corkscrewed!
Do you play on spread widening? Didn't you used to say that it was risky?
Only now I realised that I must have misunderstood the question!
For intermarket spreads (ZB-ZN, fuel oil, gold-silver, etc.), the risk is not critical - a seasonal bet on the convergence or divergence.
A slightly different matter is the so-called calendar spreads, i.e. spreads of different contracts of the same instrument. (E.g. KCU1-KCZ1, ZSU1-ZSX1 etc.).
Here it is very possible that the risk of seasonal selling spread is different from the risk of seasonal buying spread. Depending on the current state of prices (kotango-backwardation).
But I (due to my modest knowledge so far) refrain from making categorical statements. Therefore, I can safely say - that I have not previously stated that in the general case - "it is risky to play on the spread widening"!
In principle, you could write your own seasonality indicator. At one time I even started to do it. It looks like this: We take an accumulative graph of asset yield of Ln(Price(t)/Price(t)-1) type, cut it into equal pieces of one year length, then each of them is detrended by a linear trend (linear trend is appropriate in this case, because we will be interested in its deviation). Then the hypothesis is made that the linear approximation shows the general, annual trend, which does not depend on seasonal factors, and any deviation from it is seasonality. If indeed deviations from the general trend take place, then we will observe the effect of increasing deviation, which in the end will exceed the standard error of the sample. This will testify in favour of seasonal correlations. If on the contrary, seasonal fluctuations will be irregular (for example, half of January price went up and the other half went down), then eventually all these deviations will annihulate each other (or converge to zero relative to the general trend), and we will get slight "seasonality" disturbance not exceeding the sampling error.
Further, it is also simple. If seasonal correlations are confirmed, a simple correlation model is written, which traces the relationship between the current price and seasonal trends on a limited data window. When the correlation increases, it will be a clear sign that the current price "follows the seasonal component", on the contrary, low correlation percentage will mean that the current price behavior does not take seasonal trends into account.
Also it seems interesting to study the seasonal volatility. The methods for obtaining this information are the same. If we know with a certain probability that volatility increases or decreases in certain periods, we can make a lot of money with put/call options.
The indicator is ready!
I can't wait for my indicator to be tested in Code Base. So I decided to post it directly on the forum:
Description:
The indicator plots the spread line (the virtual equity line) of one or several instruments and combines the annual trends on the chart.
The figure below shows the result of the soybean meal vs. soybean oil indicator:
Fig.1 Multiyear spread trends BUY ZM - SELL ZL
I would like to point out that the indicator is intended primarily for working on a daily timeframe. However, if there is not enough data for a deeper analysis on the history of the daily timeframe, the indicator will take the prices from higher timeframes, interpolating them into the current one.
Parameters:
Formula is a string parameter in which to enter the spread formula. For example, if AA+B-CC=1^1^2 means that if you buy instrument AA with 1.00 lot volume and BB with 1.00 lot volume and sell asset CC with 2.00 lot volume, funds in the account will change the same way as the black bold line. If you don't specify the deal volumes, as in the example on the picture, the indicator will display the results of equilibrium deals with the volume of 1 lot.
Future- the number of candlesticks in the future, which will display multi-year trends. Shift of historical lines.
Past- number of working candlesticks before the current one.
BaseSymbol- number of basic symbol, by which the profit points are calculated. If BaseSymbol = 0, the indicator calculates the profit in points of the first instrument.
Compress- if enabled, this indicator registers the past years in the current spread. This mode is convenient for visual determination of the day when it is better to enter the position, and when it is better to exit from it. However, you can't measure the profit of previous years with a mouse in the "crosshair" mode.
Visual- displaying of information about profits for a specified period of time for each year separately in the left upper corner of the indicator chart. The period is specified by moving two vertical lines on the chart of the current symbol. In Figure 1 these lines indicate the interval from December 10 to 29. The chart shows that if you buy the spread in this period, you can (most likely) make profit. The layout by year shows that 3 of the past 10 years were unprofitable. Each year is coloured with the colour of the line it is involved in first.
You can update the calculation after moving the vertical lines by re-initialising the indicator. To initialize the indicator you can call the indicator properties window (Ctrl+I) and press OK, or simply change the timeframe to any and go back.
All 8 buffers are used in the indicator. Zero - always shows the current spread, and the 7 others - the averaged spreads for several years. The averaging intervals are set in theInterval_ variables, separated by a hyphen. For example, "7-1" means that the averaging will be performed from year 7 to year 1 inclusive. That is, for 2012, the first year of old age is 2011 and the seventh year is 2005.
Parameters:
Formula is a string parameter in which to enter the spread formula. For example, if AA+B-CC=1^1^2 means that if you buy instrument AA with 1.00 lot volume and BB with 1.00 lot volume and sell asset CC with 2.00 lot volume, funds in the account will change the same way as the black bold line. If you don't specify the deal volumes, as in the example on the picture, the indicator will display the results of equilibrium deals with the volume of 1 lot.
Future- the number of candlesticks in the future, which will display multi-year trends. Shift of historical lines.
Past- number of working candlesticks before the current one.
BaseSymbol- number of basic symbol, by which the profit points are calculated. If BaseSymbol = 0, the indicator calculates the profit in points of the first instrument.
Compress- if enabled, this indicator registers the past years in the current spread. This mode is convenient for visual determination of the day when it is better to enter the position, and when it is better to exit from it. However, you can't measure the profit of the previous years with a mouse in "crosshair" mode.
Visual- displaying of information about profits for a specified period of time for each year separately in the left upper corner of the indicator chart. The period is specified by moving two vertical lines on the chart of the current symbol. In Figure 1 these lines indicate the interval from December 10 to 29. The chart shows that if you buy the spread in this period, you can (most likely) make profit. The layout by year shows that 3 of the past 10 years were unprofitable. Each year is coloured with the colour of the line it is involved in first.
You can update the calculation after moving the vertical lines by re-initialising the indicator. To initialize the indicator you can call the indicator properties window (Ctrl+I) and press OK, or simply change the timeframe to any and go back.
All 8 buffers are used in the indicator. Zero - always shows the current spread, and the 7 others - the averaged spreads for several years. The averaging intervals are set in theInterval_ variables, separated by a hyphen. For example, "7-1" means that the averaging will be performed from year 7 to year 1 inclusive. That is, for 2012, the first year of old age is 2011 and the seventh year is 2005.
Thank you, I will have a look...