Although a change in trends or shift into “oversold” or “overbought” conditions, especially on a monthly timeframe should result in some drastic shifts in your portfolio, there are times between trends when individual positions are up or down, or when certain changes in fundamentals should impact your weightings, and there are times when you overlook something and just make irrational decisions and need to go back and tone down your excitement. Evaluating your positions from time to time is important. I think it makes more sense to look at your allocations and your expectations of return, compare it to the current trend and change positions. If you get caught up looking at which positions are down or up, and how much, you may be more prone to make a mistake either chasing your winners, or cutting them too early and either adding onto your losers too aggressively, or buying high and selling low. You need to distance yourself from past results, and perhaps for some, determining the correct allocation ahead of time based on your positions (before even looking at your current allocation) ahead of time, will be an even better choice. Don’t make a mistake worse by not correcting it, or even make it worse by actually over compensating. Now if your allocations are within a percentage point, it may not be worth the fees of correcting it depending on how much money you have at risk, and the cost of the fees. Another good time to evaluate positions other than changes in conditions and trends, are when an arbitrage deal has gone through, freeing up cash, and reducing exposure to “neutral”.
You want to both look at positions individually, as well as group similar positions together such as you make a “risk on”, a “risk off” and a “neutral” section. The particular allocation towards or away from risk depends on trend, while the individual positions depends on either value or perhaps how oversold it has become and is partially about reducing correlation. You have to weigh multiple factors but once you decide your “risk” you then need to look at how you can reduce positions in one area first, and then add in the other later. You may also want to factor in the leverage of ETFs and how that might increase your position in an area.
When you do this, you may find certain arbitrage positions that yield a lower return have a higher position size. This is likely because of change s in prices. If the deal is more likely to go through or less likely to decline if it doesn’t, it is okay to own more of an arbitrage deal not expected to go through. However, you can often trade a small handful of these arbitrage positions to roll your positions over more aggressively into the one yielding a higher compounded annual return. You also may find too much exposure in a particular area. You might have too much “risk on” in a downtrend because you grabbed a certain ETF that wasn’t strongly correlated in a time when it was oversold and it may have recovered. You may have went overboard in a particular purchase. You may have lost money in a particular area. Either way, you will want to do your best to reduce or sell position sizes, and potentially add on to other ones to come up with the right allocation for the situation. You still want to look at price, but not in the way where you look at where you bought it. If your stock was oversold and has traded well out of oversold territory, then there is less of a reason to own as high of a percentage, even when you started. So you may have owned 5% of assets in DBA, and it may have traded well above oversold and you now own 6%. You do not want to reduce it to 5%, but provided the trend hasn’t changed from the increase of price, reduce it even lower as the conditions are pa little less bullish, and when it reaches overbought it will as well. Having an idea of what a particular asset is worth is useful so you can compare relative value, but in some situations you might just be owning it because you need exposure elsewhere, you need to reduce correlation and it has reached extreme positions and seems like a more reasonable place to put your money than stocks. If you cannot think of a good reason to own it anymore, reduce or close the position out.
I like to put my positions on an excel spreadsheet, the percentage ownership the percentage of “risk on, risk off, arbitrage” and perhaps even the percentage of bonds vs currency within “risk off” the percentage of “stocks vs natural resources” in “risk on” and if I am doing a “neutral” strategy other than arbitrage I might put that down expressed as a percentage. I will look at my “rule of thumb” chart where I am prepared to rebalance my portfolio and change the allocations based on conditions and verify the market is in those conditions and try to ask myself if aiming for that seems to make sense, and to do so. I also will just add up “risk on” and “risk off” and divide the individual area by the sum of the percentage allocation of all including “neutral” to get a percentage ignoring the “neutral” that adds up to 100%. At a minimum, I should have 50% of each, but generally will aim to support the information given the trend. In a downtrend, I should have at least 50% exposure to “risk off”, but preferably more. Sometimes it may make sense to own 3x etfs if I notice a lot of good arbitrage deals so I can reduce the exposure to arbitrage by increasing my “risk off” and “risk on” exposure through leveraged etfs without going on or adding margin.
If this works for you, great. If not, fine. The interesting thing that I will do is although I have these “aims” I will sometimes increase margin to take advantage of an opportunity, and occasionally will add additional aggression based upon shorter term signals if they are ideal. If I think we have sold off enough, I will not usually act, but if it continues to further extremes, then I will act decisively and aggressively. Perhaps those who trade the PPT can look at the hybrid score and buy the SPY or even TNA when the market reaches “oversold” and give yourself room to add more if it declines lower. You do not have to depend entirely on the PPT, but instead can use it when better than usual extremes are reached. So if 2.59 is oversold, perhaps you only buy 2.45. The stability of your other positions will allow you to add some leverage to place the trade with plenty of room to add additional leverage, without as much volatility if you are wrong.
How well does this “system” work? over 10% during a 2 month period in my testing. Only recently have I been more liberal with leverage due to watching the changes in the portfolio failing to move significantly downward from any time. It was more like 7% until recently, but a deal that is pending and has froze any transactions at this time, and expected to close before the end of the year will add an additional 1%. You could be more aggressive with it, or more conservative. A 5% per month return for a year is nearly an 80% return by year end.
In my experience with the PPT, I believe using TNA or in the money calls with lots of time on it is probably the best way to play it. I tried out of the money calls with minimal time on it with mixed results, sometimes I would nail one out of the park, other times, the position would either expire, or had to be sold well before it did. Focusing solely on the PPT and using calls wwas far too aggressive, even with a large amount of cash on the sides as I had. So I then was forced to go back to the drawing board. So I started paper trading first, which is what I am doing now, and this way I will have something to compare with when I use the same methods, only with a more aggressive rebalancing using the PPT, within this system. I don’t believe I want to use TNA just yet, and instead will just be using SPY and other non leveraged ETFs without the PPT. With the PPT, I am not sure I want to go there either. I will be using margin on arbitrage and occasionally elsewhere, so adding leveraged ETFs while adding margin during oversold conditions even with the PPT accuracy seems a little bit more aggressive than I need right now.
When I went back to the drawing board, I had forced myself to look at the “kelly criterion” from a different angle. Previously I would stick to one method and try to leverage it to the max, aiming for a “full kelly” or “one half kelly”. I realized that multiple bets independent of each other that in total added up to a full kelly or half kelly was a superior use of capital and supported by the same criterion, and understanding the nature of volatility and portfolio growth actually shows the hidden lessons. Risk less with more positions that seek independence of each other. I was only starting to pick up on this idea intuitively as I have only been trading since 2005 or 2006.
Additionally, it is more prudent to aggressively buy the oversold than to sell the overbought.You can certainly prepare yourself to add even more if it reaches a more extreme reading or a second closing “oversold”.
If we want to get into more specifics of how to do this, then you want your positions within the risk profile of the current trend to be weighted based on expected annualized growth rate per year. Using the <a href=”http://www.cisiova.com/betsizing.asp”>kelly criterion calculator</a> to factor in downside risk, probability, etc and then compounding that growth rate over the amount of times you will trade in a year will help you determine what kind of return you will get if you were to bet the full kelly. If you get a “kelly” of 100%, recognize that the calculator does not go above this, so the return per risk (kelly is used) will actually be higher since you are not getting a measurement of full return since your risk is capped at 100% invested. However in all other cases, you are getting a return that is put in the proper context for which it can be compared, as soon as you figure out how many of these deals you are able to do within a year. If your kelly criterion is 75% and growth rate is 5% and you do 1 per every 2 months or 6 per year, you can take the number 1 and multiply it by 1.05 and then the resulting number by 1.05 6 times to get 1.340096 or a return of 34% per year, if you were fully invested. If another asset you project based on the probability and so on you calculate a growth rate of 2% per trade but can trade it 12 times per year, you will multiply 1.02 by 1 12 times and get 1.268 or 26.8% per year. Both of these trades should be made, however the weighting of the return per kelly per year that yields 34% should be larger, while the 26.8% trade should be smaller. You may want to add up the return of all your options and divide the individual return by that total to get your “weightings” for that particular area. Or you may perfer more aggressively weighting to the one expected to yield more. I am not aware of how to figure out the “ideal” amount to consider the best combination of investing in the best return, as well as creating enough risked in the independent bets to reduce correlation and give you the best reward on the least risk possible for your overall portfolio, but since you are making a lot of assumptions here based on the probabilities of a deal going through or a stock returning that much over a given time, any way to determine this would be based upon PERFECT knowledge of those probabilities, which is not something you have in investing. Therefore just some attempt to weight towards value while still staying invested in areas that reduce correlation will be productive, but finding the perfect balance is more of an art form and with values constantly changing, and cost of fees and other risks to consider, and considering how one particular risk correlates with the other it is not something you simply can master with a mathematical formula. However, being aware of this stuff and knowing what factors play into the formula, were information perfect, and the variables you need to consider when making decision, can make this “art” much more in alignment, so the “science” of investing certainly is important.
This is the purpose between my long posts of explaining the kelly criterion and the correlation reducing and boosting return. The trend report is only going to tell you about the conditions that exist. In order to use it in a productive way I explained how you can use it. I created the “trend Trader” post as an example of what someone might do with it. If the probabilities of each trade were perfectly known and you could predict the results and such, I probably would look more into the exactly perfect allocation. I believe it would be dividing the percentage of “kelly” evenly across all assets, but that would assume the time frame of all trades were the same and would never exceed 100%, which is sometimes flawed. If a coinflip paid 10:1 and another paid 9:1 you still have reason to invest in both and you want to invest more in the 10:1, but the formula accounts for that. The 10:1 return would give you a kelly percentage of 45% of your capital, the 9:1 would give you a kelly percentage of 44.1% of your capital. If you had 3 profitable situations to bet on you probably would bet 1/3rd the kelly across all 3 which would be slightly higher in the 10:1. If not for the fact that you can lose more than you put into it and that you have to pay back the money by a certain time frame, you could justify always betting the full kelly across as many as possible, and then when you owe money because you lose simply continue to bet on these profitable situations. however, that is not realistic and this is yet another variable which makes investing an “art”, even if everything was more exact. So we just have to accept the principals of the kelly and understand the context of how we are using it, and do the best we can. However we certainly can use it to compare return on risk among all our investments we view as profitable enough to act on, and use that information to assess which investments should be weighted more heavily.
I personally believe that you want to first break your investments up by “risk profile” categorize it as “risk on” (tends to go up when market goes up) “risk off (tends to go down when market goes down) and “neutral” (tends to have no correlation of market direction or very little). You should assign your investment risk among those 3 groups first based on trends and contrarian indicators and such, and then within those categories you can compare. The reason is, you still want your portfolio to work WITH the market rather than against it, and it is difficult to evaluate “return” one an arbitrage deal the same way you do for a return on a value stock, or a return on bonds or currency. You still need the investment in bonds for your overall portfolio, and unlike in “theory” where we talk about coinflips, there is no such thing as an independent bet in investing that is not correlated with another bet, when the bests are made at the same time, or overlap in time frame.
After you break your investments up, then you can compare the kelly criterion based on the probabilities you assess of each arbitrage going through, of each investment going the way you expect, and of the bond yielding the result it says without default, or the probability of a bond ETF or currency ETF doing well over the time frame you trade it. And of course you also factor in the loss expressed in a negative (on the kelly criterion calculator) if the deal doesn’t go through, if the investment doesn’t go well, or if the nation defaults or hyperinflates and the probability of it doing so. Then you multiply the return as discussed before, and you compare them side by side. You may have to use your own judgement if there is a deal that has more uncertainty about it’s return, or greater certainty about the percentages expressed, or factor in things yourself such as adjusting your portfolio if it is too correlated towards a specific event, but that will give you the returns to make sure in most cases that a 20% annualized growth rate (including risk) has more invested in it than a similar deal with a 10% annualized growth rate.
As you do the work of this several times, it can become a bit more intuitive, and you can start to consider making it a bit more complicated by breaking things down into more than just “risk off” and “risk on” and “neutral”. Instead you can add more in the “risk on” area to achieve roughly the same portfolio. What I mean is if you have multiple minorly correlated bets in that area, you can increase the percentage “risk on” because your actual “risk” may be mitigated by holding assets that may only be slightly correlated with the S&P and in a much better position to do well. You can also have “value vs growth”, and you can learn to handle a bit more at once. you can have more in “neutral” if you understand the risks and can assess it better, and you can also consider other “neutral” investments. You can consider adding SHORT positions in either of the 3 areas to offset the allocations as well and this way you can bet against those overvalued allowing you to bet more aggressively on specific areas which you feel are undervalued. Much of this would be harmful if you didn’t have an understanding of all of this. Much of it would be harmful if you weren’t comfortable with it. So it can be a lot to take in at once, which is why I believe learning how to make the exact calculations based on probabilities that you asses. When you start you also will have to leave yourself a greater margin of error when guessing the “probabilities” of events, and invest more in “risk off” (specifically cash) than elsewhere. However, you can be a master investor by maximizing this principal of reducing correlation of your portfolio, weighting towards value, adjusting towards trends including overbought and oversold, adjusting on weekly basis, and perhaps even adjusting on daily extremes and putting enough capital towards “neutral” positions, as well as “risk on” and “risk off” to give yourself the ability to profit in all conditions. If you start with 50% cash 50% stock and rebalance you can even beat a random market by maintaining 50%, but if you add in arbitrage that profits in all conditions, trend following which positions you better early on in the trend to by weighted to take advantage of the trend, as well as using extreme conditions as well to anticipate a change in trend, and the principals of comparing value relative to each other and evaluating your positions and repositioning towards value as the value changes, (as well as using principals of contrarianism at times) while evaluating the return over time for arbitrage plays and other neutral plays and evaluating and repositioning based on annualized returns after factoring in risk, you will have a great chance of being able to have crushing returns. You may wish to add in growth and momentum trading as well in the “sweet spots” of a growing company, but you really should try to stick to what you can understand with in the context of this structure. If you don’t understand arbitrage, but understand delta neutral option plays, or understand pre earnings, or pair trades, go for that. If you don’t understand value investing or don’t have the patience, fine. If you prefer to scalp trades in each area, you can still be positioned according to the greater trends to a more limited extent, and use very short term swings in the market to follow those trends… then go for it. If you don’t have an edge and worry about currencies or bond trading, stick to cash. The principals of risk management when structured in the right context are fairly universal.
Recent Comments