Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
... since I live almost entirely off my investments, I'm always interested in high yield investments. They are also the source of my biggest investment losses...
Please excuse me rudely jumping in on your conversation aim hier, but you might be interested in the High Yield Portfolio discussions over on the UK's The Motley Fool.
A reasonably starting place is http://www.fool.co.uk/valueinvesting/2006/vi061011.htm (you may need to register a user id but its free). There are many articles about the strategy and there's also a discussion board to read through.
I'm personally effectively 100% loaded in a similar strategy, but have extended the concept to include a modestly sized element of leverage as well - my speculative component. Generally therefore if the HYP paces the market average (so far its comfortably outpaced the index), and the leverage play(s) outpace the cost of debt/trading costs, then its an overall market average beating blend over the longer term.
Until recently I had more or less been constantly leveraging all of the time, but have since decided to switch to a timing based approach, using AIM amongst other indicators (Tom's iWave, Fosback's seasonal, Dow Theory, Technical Indicators etc.).
The intent being that any large price declines will generally be preceded with a high risk indicator level such that the leverage wont kick-in until after the largest price declines had already been encountered. Then leverage from the low-risk indicator until a return to high-risk occurs at which time close out and the gains serve as income supplement bonus.
Having moved out of the rat-race some 4 years ago (single parent with two aged 10 and 13), I too am in a similar position to yourself and live principally off investment income, so I know exactly where your coming from with respect to seeking out yield and an element of inflation beating growth.
Regards. Clive.
This switching type strategy is along the lines of what I was personally intending moving towards, except my personal preference is holding 100% constantly in equities and complimenting that with periodic modest levels of leverage during low iWave periods. So if the net investment benefit against borrowed funds during leverage periods exceeds the cost of debt service and trading fees/spreads, a net out-performance of the underline is apparent.
My thoughts had initially been towards using variable levels of leverage according to the iWave level - a sort of AIM on the iWave. Subsequently however I'm more swayed towards using a red/amber/green type indicator on each of a range of measures of which AIM and iWave are just two, and proportion leverage exposure according to the percentage green indicators compared to the total number of indicators. For example if my maximum leverage is 50% of the fund size, then with 20 indicators in total and 10 green cases then the actual leverage level would be 25% (or alternatively perhaps discount the number of greens by the number of reds). This entails having many different types of measures such as relative strength, technical charts, Bollinger bands, moving averages etc. each with their own red/amber/green flag, which whilst initially appearing to be a considerable effort, is in practice a relatively quickly performed task.
On a separate note, one striking feature of Tom's Fraternity account is that most of the differential arises out of events in the first six months in 2002 when a move out of stocks preceded the subsequent stock price decline. Since then the two appear to have moved more or less in step with each other.
Regards. Clive.
I see that you've added an animated icon to aim-users.com now Tom.
Only seems to show up in Firefox though, or maybe thats because the version of IE I occasionally use is a bit dated now.
Regards. Clive.
Tom, here's some useful details that you may be interested in of how to create and install your own icon in browser address bars and favourites list.
http://www.chami.com/tips/Internet/110599I.html
Here's another link of how to get your web pages speaking
http://vhost.oddcast.com/vhost_minisite/sitepal_tour/index.html
Sorry guys, better end this thread. Might give some the needle.
As they say Tom, bottom fishers become cotton pickers.
After having posted my last message, I was just about to close the browser tab for yahoo and noticed that they have growth down for 100%+ next year.
Surely I thought that's got to be an over-estimate and such not a worthy stock. BUT
http://investor.unifi.com/phoenix.zhtml?c=82594&p=irol-newsArticle&ID=921570&highlight=
reveals how they're expanding and expect such to produce positive cash-flow.
Released 25th October and around that day the stock jumped from 2.4 to 2.8 odd (17% odd).
http://finance.yahoo.com/q/bc?s=UFI&t=3m&l=on&z=m&q=l&c=
But has since dropped back to 2.56 odd (6% up from the 2.4 level).
As such it looks like they've been attempting to turn around, reduced costs, improved efficiencies, expanded and as such might be a reasonable candidate for recovery. With the analyst looking to 108% growth in 2007 however, a current 133m market capitalisation, 0.37 Price to Book (inferring 400m book-value), and acquisition of 65m, it does seem a little short on the growth side. The stocks lacklustre performance appears to reflect that.
Their report indicates $1B of property, plant and equipment assets. $250M debt, such that at a $133M market cap, $383 shareholder equity and 0.37 Price to Book would mean buying a stake in those assets at a discount. Potentially a buyout even at 50 cents on the dollar would yield a 50% gain.
Anyone got a spare $60M that could lend me for a few months? - Come on Tom, I know you're quite heavy in cash at the moment ;>) If so, does anyone want to buy an old loom?
Clive.
Hi MM,
What I personally tend to do is something like check out the figures on finance.yahoo.com for example
http://finance.yahoo.com/q/bs?s=UFI shows that the decline in book-value is somewhat offset by the decline in debt.
Another source is to go to the horse mouth, go to the company web page and look for something like investor relations or failing that do a Google on the company name and 'investors' and track down the latest online report, usually in PDF format. When reading through the report usually I scan the directors outline near the start, and then read through the report backwards (UK reports at least tend to show the facts/figures at the end). For example
http://investor.unifi.com/phoenix.zhtml?c=82594&p=irol-irhome
- in this case (from a very brief scan) in the directors summary they outline how the textile industry in the US is and is likely to continue to be threatened/decay by cheap labour countries....blah blah blah. But they do outline how they're branching out, lowering costs attempting to turn around etc and an inspection of
http://finance.yahoo.com/q/ae?s=UFI
shows some potential of turning earnings around from negative to positive. But bear in mind I believe its only a single analyst's view in this case.
Typically price rebounds occur due to unexpected better results/performance, so if they exceed expectations then the price will rise, if they just meet or fail to meet expectations the price will likely just stay level or fall. So to gain from the stock, you have to believe that they will do better than expected.
In effect its a play as to whether you believe the analyst(s) assessments are over or under done and if you are correct in the belief that they've under-estimated, then there's potential to profit.
Regards. Clive.
Thanks Tom and Aimster.
Whilst it does protect one from, as you said, "throwing good money after bad", it's really using only 1/2 of the AIM process. As such, it's not really AIM any more, per se.
The released funds would have to be invested elsewhere at some time however, and maybe a market wide AIM (or even one based on individual stocks), could be used to identify possible timing and selections for such reinvestments. So not necessarily only half the picture.
In the 2nd edition of Mr L's book, around pages 175-178, he does propose one option of buying just a single stock at the offset, working the sells but not acting on any of the buy signals - instead buying an alternative stock with the proceeds, so as to provide initial above average volatility, but build up diversity in holdings over time.
Much of AIM's individual signals are likely to be a consequence of the tidal movement in the overall market, so selling one stock and immediately buying another might not be the best choice, especially if the market wide AIM measure indicated prices being more closer to the next sell signal than buy signal. The benefits secured from the sale, as Tom suggests, might provide a time-window period in which to find a suitable reinvestment choice. In so doing the associate buy signal of the other half of the AIM component is not restricted to the same stock, but instead across the entire market constitute set and as such is likely to be encountered more quickly.
Taken to an extreme, we could run a virtual AIM on every stock, or rather AIM suitable stock and identify a candidate that was indicating a buy of comparable size to the sell signal from our actual holding and rotate the sell order into that stock, thereby completing the AIM buy/sell cycle, but in near instantaneous time. Not only would this increase the overall rate of buy/sell rotations and hence benefits, but it would also tend towards having equal numbers of both buys and sells over time rather than perhaps a biased number of one or the other.
Thanks again guys, you've given me more food for thought on this.
Regards. Clive.
I like to buy above average yielding stocks - simply because I like/need the income. A problem with this is that when the stocks price subsequently rises, it leaves you with a dilemma of whether to continue holding that stock or switching out and into an alternative holding. As AIM generally selects reasonably good times to buy and sell, it seems reasonable to leave that decision to AIM alone. But I don't particularly want to add to a holding once having made the initial purchase because if the stocks high yield is an indicator of bad times to come, then I don't want to be throwing good money after bad.
So I set up a test using the following AIM values.
Initial Cash 90% (e.g. 10% invested into that single stock - being representative of an intent to hold five or so total stocks individually, each being AIM'd in a similar manner using a common cash pool (total 50% of fund value spread across 5 stocks)).
Buy Safe 1000000% (so AIM never signals a buy)
Sell Safe 10%
When run against the UK's monthly based FT100 Index price data (full history since its inception in 1984) it produced substantially higher ROCAR over that of B&H.
Reversing the prices as though index had gone through time backwards, then the B&H was down -82% to a 0.18 factor of its original value whilst this approach was down -8% or so.
The high ROCAR for the upward price motion case is indicative that AIM does a reasonable job of identifying good sell prices. The lower loss on the downward price motion side confirms this.
One further slight tweak I did find helped improve results was to adjust Portfolio Control over time by a factor of the previous periods price divided by the current periods price, but only when no trade is made in that period (being lazy I didn't adjust the Excel cell formula for both, only the ELSE statement). So doing in effect increases PC when its a down price month, thereby reducing the prospect of a sell signal, and lowers PC in a price up month, increasing the prospect of a sell signal.
Has/does anyone else used such a sell only based AIM?
Regards. Clive.
Yes Tom, the UK's had a couple of good years now. Personally I've seen around 28% gain in 05 and around 14% so far this year.
Interest rates look a cert to rise this Thursday to 5% from 4.75, but overall the markets not too hot, I estimate only around 6% to 7% above fair price levels.
There is some talk about dequity however. Pension funds moving out of the bond/equity typical balance into a wider spread of asset classes, so 2007 could turn out to be a good year to be holding a large cash holding.
On further reading through SGAM I wasn't too impressed. Looks like they hold back 50% of the fund value at the start of each month and use derrivatives to leverage up to an overall 200%, which means something like 400% scaling. Whilst they effectively guarantee that 50% of the fund will remain intact in any one month, it wouldn't take too much to knock the fund into an unrecoverable condition.
Regards. Clive.
Tom,
I'd be interested to see the iWave raw data that is shown as being available from a link off the bottom of iw.htm, but leads to a 404 not found error.
Is the data still available?
Regards. Clive.
Hi Toofuzzy,
If instead of using a Vealie you were going to pull out some amount of excess cash, I would use it to start a NEW AIM account, not put it an existing one.
Personally I wouldn't start a new AIM with just 5% or so of the previous AIM'd stocks account value (assuming 50:50 cash/equity split and 10% min trade size), unless it was a very large holding.
Unless perhaps with the intent that when a subsequent sell signal was raised, then that entire holding was closed out, perhaps with the intent of moving all proceeds back into the original holding or another account that had a relative close price level to its target accumulate trigger price.
Regards. Clive.
A Vealie question.
I've read through various historical posts and found some references of opposing views of Vealie's. It would appear that some view the concept as not building up the cash reserves in the by the book AIM manner in anticipation of later discount price levels, whilst others highlight the missed benefits of being more out of stock when prices are rising and having too much cash reserve.
Would actually trading all AIM signals but moving the released funds from sell signals into more attractively priced stocks such that the overall fund value across all AIM positions was maintained in line with that indicated by iWave be a viable midway solution?
I appreciate that there's the trading costs element involved, but if one AIM account was relatively close to a buy compared to another than was indicating a sell - but a Vealie was being considered in view of having too much cash reserves overall, then it appears to me that rotating the indicated sell amount of apparently overpriced stock into the one that more relatively underpriced would be the more appropriate choice.
Regards. Clive.
Hi TooFuzzy There is also the issue of "why do I want to save and make more money"
Play golf
Spend more time with my family
Avoid my family
Work less
etc
You can AIM by spending 1/2 hr per month. (The Quick Aim Calculator helps alot in that)
Anything else is taking away from the reason you want to save in the first place.
That much time! ;>)
Part of the reason I toy concepts and ideas is both because I often have too much free time as my trading activity is most likely a lot less than your own and secondly because its a hobby sort of thing.
I already have/partake in all of the above, except I cant work less as that's already at zero.
Regards. Clive.
Hi Tom, I have a proprietary model similar to what you're discussing. It consists of 10 separate business sector ETFs plus three smaller cap value and growth index ETFs.
Thanks for the information.
The 13sector.htm link no longer seems to be accessible off the aim-users site now Tom.
I've also noticed that the iwdata.htm link at the bottom of the iWave description (iw.htm) page is also broken.
I personally prefer the idea of the cherry-pick approach of holding a number of stocks in each sector as the price volatility is generally higher.
Regards. Clive.
I've been considering the use of AIM on sector type ETF's, but it strikes me that this might not be an appropriate tact.
The more AIM trades in and out of stock, the greater the rewards tend to be. By limiting to sector ETF's that comprise a number of stocks in each such sector fund, the oscillations in price of the individual stocks will tend to cancel. That is the individual stocks might be anticipated to have a higher average high to low price range (in percentage terms) than that of the sector fund. A similar situation occurs for the Index - its high to low yearly range is typically lower in percentage terms than that of individual stocks. By replicating an ETF, or partial emulation by holding individual collections of stocks within the sector then we likely trap more price range oscillations.
Which led me onto another thought train. If instead of using AIM on stock price, what about using AIM on relative price instead. So for each stock we'd record the index and sector index figures at the time of a stock purchase and weekly, monthly... whatever, record the relative price performance of that stock compared to the index and relative price performance relative to the sector.
Where a stocks relative price performance outpaced the sector and/or index, then we should be considering selling some of that holding and when a stocks relative price performance under-performed the sector or stock then we'd buy some more of that stock. In turn we could also measure the relative price performance between individual sectors and perhaps increase or decrease the exposure between such sectors accordingly.
Ideally for this to be practical, we should hold a number of stocks within each sector. Historically this has been my preference anyway, typically holding 3 or 4 stocks in each of 4 or 5 sectors. The principle reason is that it lowers survivability risk - an example being if a sector comprised of three stocks each equally priced at say 100 and each having comparable earnings and profits, then if one of those goes broke, the other two pick up on that failed companies assets and business and typically see both survivors increasing their profits/earnings and stock price by 50% apiece. An oversimplified example I know, but hopefully you get my drift. This is important because AIM has a tendency to overweight failing stocks, but sells out from strongly rising stocks. Conventionally for buy and hold this isn't an issue as the small number of failures is offset by a comparable small number of stock performing particularly well.
It strikes me that potentially such a relative price performance based measure over that of just price alone is a relatively simple means to use AIM to generate a larger number of rotational type trades across stock, sector and market wide levels - in a more profitable manner than had just sector based ETF's AIM's been used alone. That is rather than most round-robin buy/sell type benefits be achieved locally at the stock level, round-trip benefits might be spread between a stock and sector, stock and market, sector and market, sector and stock ... etc.
Any thoughts/views?
Regards. Clive.
Hi Ken.
The buy and hold turned around a 10.5% p.a. average profit.
The AIM with around 50% average exposure produced a 7% average.
Both excluding dividends. RYNVX didn't pay that much in dividends over the years, the best being in around 2005 when 3%, otherwise pretty low on that front.
Assuming a cash-deposit rate of 5% over the same period, a 50:50 buy and hold and cash would have produced 7.75% (half of 10.5 + half of 5). Your figures indicate (I'm assuming excluding cash deposit benefits) a 7% return, add on 2.5% for the cash interest and that's 9.5%
As I see it, AIM picks off relative highs and lows, betting a relatively small amount on each such case, for example with minimum trade size of 10% and 50% exposure, that's an amount of around 5% of the overall allocation to that stock. A swing between high and low effectively results in a round trip of 5% stock being added/withdrawn. Similarly the opposite case could arise e.g. a swing high to low for a sell/buy rotation respectively. If the gap is 20%, then the total benefit as a proportion of the overall allocation to that holding (cash and equity) is 20% of 5% = 1%. Generally the more of these round trips the better as each effectively lowers the average price paid for the stock. Where a trend is sustained however, if the trend is up, stock is sold off - profits taken. If the trend is down then more is added - cost averaging the overall price downwards. I would have expected with an average of over 6 matched buys and sells p.a. overall, that the benefits would have been much higher. So something would appear amiss.
The graph for RYNVX
http://finance.yahoo.com/q/bc?s=RYNVX&t=my&l=off&z=m&q=l&c=%5EGSPC,%5EDJI
shows the effects of the funds leverage basis (150% beta in this case). The 200% run up in the underline produced a 300% comparable return for RYNVX, to around mid 2000 that was. Subsequently the index fell 50%, but RYNVX halved and then halved again, if not more, to around 2003 when AIM ran out of cash. Following that, it has performed comparably to the underline. End to end its generally paced the underline.
Having gone effectively heavily in from that low in 2003 (ran out of cash), it ended with a near reversal with AIM being near fully loaded in cash at the end date.
Overall therefore it would appear that this would have been a lousy choice of holding, however in view of such, in comparison to perhaps 11% total benefit buy and hold, 7.75% for split 50:50 buy and hold and cash, AIM's 9.5% against around 50:50 cash:stock split isn't to poor.
I'd guess overall therefore that it was the half and half again motion between 2000 and 2003 odd that was the crippling period for AIM. Loading up with more and more stock with ever decreasing price levels, the averaging down benefits just didn't work out too well - as such the application of averaging-down were a relatively high overall average, such that its only towards the end date did the stocks price has reverted to around that average price of stock level.
should we consider 7% Aim by the book a good result?
My guess is that in view of this being a good example of a bad AIM candidate, that the end results weren't too depressing, bearing in mind that other holdings that might have been good examples of good AIM candidates would likely have faired far better, such that an overall combination of such holdings produce satisfactory results.
Comparisons of performance are always relative. For example between the start date and 2000 odd, the buy and hold crowd would have been blowing raspberries at the AIM crowd in reflection of a 300% rise in a handful of years. Between 2000 and 2003 the B&H crowd would have been weeping. End to end the tortoise and hare have finished together. Overall however AIM likely had far less volatility in paper profits/losses along the way, and its mechanical nature would have been more likely to keep an investor acting rationally, buying at apparent lows and selling at apparent highs, rather than perhaps dumping out at exactly the wrong time (buying high and selling low).
Just my opinion of course, bearing in mind I don't even use AIM (but I am looking more increasingly likely to do so over coming years).
Regards. Clive.
Hi Firebird,
Your example data can be used for the basis of why diversifying and using AIM on individual selections but sharing a common cash pool can be of significant benefit.
If you had another stock that had an inverse correlation to RYNVX, such that when it ran out of cash it could have borrowed from that other account, then the final results for RYNVX are considerably higher, approaching that of the Buy-n-Hold total returns, yet ending up with far less equity exposure.
I would suspect that Tom's sector ETF's based AIM's would exhibit such potentials over the longer term.
Regards. Clive.
With $140B group size, and 20% p.a. average, 24/7 works out to $3,200,000/hour.
Must be a bit difficult to take some time out when earning that amount! Every second he spends in the can amounts to nearly $900. Puts a whole new meaning on the term Spending a penny.
Clive.
It was late when I posted #21156 (local time = board shown time + 5 hours). Too much, Beta (blocking) and not enough Omega3 ;>)
The market has a beta of 1 by definition. The reason for that is because the term beta is useless without any sort of benchmark. The proxy for the market is usually the S&P 500 since it is impossible to use the actual market which would be a combination of all assets (realty, equity, fixed income, etc).
The S&P500 for me, being UK based has currency risk and therefore diversifiable risk. As you say beta is relative. From a universal perspective however, I was inferring that in respect to what Dave posted, perhaps selecting a local market that had been relatively quiet, and selecting within that low priced assets, may, relative to global subsequent activity, produce subsequent above average rewards compared to that of having selected high priced assets from within noisy local markets.
As Dave also pointed out however, AIM's trading element benefits more during zip-zagging price ranging, attempting to pick off lows and highs along the way. The risk-free (cash) component provides a degree of downside stock price protection. As such it betters buy and hold in two out of three cases, losing out only to rising price trends. If you reversed AIM's buy and sell signals, adding to positions when AIM signalled a sell and visa-versa, then it would have a tendency to outperform during Bull phases and lower losses during Bear phases.
Such sayings as bottom fishers become cotton pickers and the trend is your friend spring to mind. AIM knows this and as such limits its trading accordingly. Collectively AIM lowers day to day ongoing volatility - that is your paper profits or losses swing less than that of buy and hold. However AIM is not a holy grail and a penalty is that over the longer term it is less likely to outpace buy and hold on a real profits across a start to end period basis.
Regards. Clive.
That would make more sense Dave.
As in the stock market as a whole has a lower beta to that of individual stocks, lower priced stocks typically have above average beta. After all the market wide beta is the average of the constitute members.
Beaten down stocks often have a more positive beta. That is if bad news has scared investors out of a stock and/or has already been priced into a stock, then further bad news has a tendency to have less of an effect the closer the stock is to book-value or below. Reversals arising from good news however typically come as a surprise and the upside beta potential is greater.
This is an interesting concept as typically predicting beta is as difficult as predicting price performance. More often its the stock that's been the quietest that subsequently throws us higher beta, whereas previously high beta stocks often fade into subsequent low beta. Possibly then historic low beta and low price are indicators that the stock is more likely to encounter positive news that lifts the price and hence the beta.
Regards. Clive.
Hi Dave,
If you want the 252-day volatility (1 year based on 252 business days), you have to multiply the daily volatility by the square root of 252. This is based on Einstein's equation (called the T to the one-half rule) to find the distance travelled by a particle in a Brownian motion!
Let's say we have two stocks one with a day's price motion of 1.0, 1.5 0.5 another with a price motion of 10, 15, 5 - rather extreme examples I know, AIM would love it if such price motions were for real ;>)
The first is simply 10% the size of the second, so both stocks moved through comparable price motions.
The standard deviation (volatility) for the first is 0.408, and for the second its 4.08.
These standard deviation figures provide an indication of the probability of the price motion, the 'Bell Curve' distribution such that in 68.26% of cases the price will fall within 1 standard deviation (mean price +/- 0.408 for the first stock, mean price +/- 4.08 for the second), the chances are in 95.44% cases it will fall within 2 standard deviations (mean +/- 0.816 and 8.16 respectively), in 99.74% of cases it will be within 3 standard deviations ... etc. Sigma and standard deviation are effectively interchangeable, hence such references to 6 sigma events - representing effectively a extremely high deviance (6 standard deviations) away from the mean.
The square root of 252 is 15.87
So we have 0.408 * 15.87 = 6.475
and 4.08 * 15.87 = 64.75 respectively.
Proportionally the same amount of motions relative to their respective sizes.
A simile might be to consider Brownian motion, but in two different universes where in one the dust particles where ten times the size of the other. Their motion is still proportional.
O'Higgins in his book 'Beating the Dow' and Fosbacks 'Stock Market Logic' claim that there is some low price higher subsequent volatility correlation.
If I recall correctly, subsequent to publication O'Higgins book the 'Dogs of the Dow' strategy became ever worse and under-performed, highlighting that it was a simply mean reverting with a deviance over the period measured but not subsequently. Similarly if you follow the low price predicts higher volatility, likely the historic 'favourable' deviance across the period which both O'Higgins and Fosback measured their low price/high volatility effect will be subsequently corrected by a reversal.
AIM strives to buy low sell high. Following such a recorded high correlation of low price to subsequent high volatility effect is somewhat comparable of attempting to buy whilst prices were high on the back of relative strength, often such followers are hit hard with reversals. Of course however the effect may persist for a period of time as does Relative Strength over some periods. The simple truth is that it may or may not work over the period in which we chose to apply the rule, but pressed to give odds, I'd favour slightly higher odds of it not working given that its just cycled through a period in which it did work.
Regards. Clive.
Having briefly scanned through the your web site Dave, one significant error that strikes me is your mis-undertanding of square root volatility.
Your claims of lower priced stock representing higher price volatility candidates is totally unfounded if you correctly interpret SRV. Your figures present constant square root time whereas they should be based on sigma square root time which as such negates the general case low price higher volatility argument.
Such a relatively simplistic mathematical error detracts from the likely quality of the Core Position Trader program and/or Consultancy Service so you might like to review that section.
Regards. Clive.
Want a bit more volatility for your AIM account, check out the
SGAM CAC40 up to 200% exposure ETF
http://www.euronext.com/trader/chartsanalysis/0,5372,1732_6847,00.html?selectedMep=1&idInstrumen...
A whopping 50% gain nearly over the last 12 months.
And check out the chart for apparent AIM trades.
Regards. Clive.
Hi Newrider, Stochastic? That is ( ( current - low ) / ( high - low ) ) * 100
Just a guess as I'm not familiar with IBD.
Hi all.
My current investment style is to be 150%, market exposed all of the time. I am however considering a move to using AIM instead of my existing buy and hold - but with periodic sector/stock rotations as and when prices appear relatively high/low (price driven rotations) approach.
My initial idea for using AIM was to simply throw 150% virtual fund value at the conventional AIM (not use AIM-LD), so a 66% iWave would mean 100% of my actual fund value being stock exposed. As such that would reduce my actual external borrowing requirements over time as much of the borrowing could be internally funded from the virtual/actual cash pool for a large proportion of the time (up to iWave levels of 66%). Of course if iWave signalled higher exposure levels, then some external debt would have to be obtained.
I'm considering using sector ETF's as I like the independent moves in individual sectors whilst collectively providing comparable market wide diversification. My thoughts are that if I note the individual ETF values at any one time, proportion them and then total that, I might assess the allocation to be in each individual ETF by using the individual AIM exposure indications. For example let's say just two ETF's are used for simplicity, one currently priced at 40 the other 60. AIM indicates 30% and 70% exposure respectively at the current time, so I need to add to the second and reduce the holdings in the first.... perhaps rather too much of a simplistic example, but hopefully you get the idea.
I might kill three birds with one stone here. Firstly I reduce or even eliminate my existing real borrowing levels whilst likely securing comparable returns over the longer term through using AIM instead. Being 'unemployed' effectively makes securing such debt that more difficult, so this is potentially a big win. The second is that I would expect to average 100% equity exposed through AIM, I should therefore secure returns in line with the internal rate of return - again a potentially big win. Lastly it somewhat automates my current manual sector rotations. Distancing myself from such decisions and allowing AIM to do the work instead is a bonus in my view as I'm all too aware of the buy high sell low type human instinct that periodically prevails.
As an example of the potential, I downloaded the AimBareS.xls from Tom's website. In that spreadsheet there's a CAER stock sample data set for July 1994 to July 1996. On a buy and hold basis, the stock returned a 24.6% p.a. capital appreciation average. AIM with its 30% initial cash, $500 min trade size and 5% Buy/10% Sell SAFE levels produced a 34.7% p.a. average return. The XIRR however was a whopping 61%.
To clarify XIRR is an Excel function that takes each individual investment amount and date and determines an overall per annum growth rate based on a series of variable injection and extraction of funds into/out of stocks over time, its a bit like what Tom uses his ROCAR for. You simply plug in the dates and amounts invested/withdrawn and out pops the result.
I appreciate that the CAER example is likely an oddity. For example taking figures from Robert Lichello's book shows that at other times XIRR returns are significantly lower, some times less than cash over extended periods of time. Typically however relatively shortly after such low benefit periods the figures turn around again, usually quite strongly. Providing there is no specific end date involved therefore, I can simply ride through such poor periods in anticipation of better times.
As I've mentioned before, reducing year on year volatility in paper gains/losses is of little interest to me, I simply intend to hold for the longer term, such that I ride through shorter term paper losses with little regard. That is such shorter term paper losses are meaningless unless they have to be turned into real losses. The increasing of the AIM's position volatility to be more in line with the wider market is not a concern for me.
I'd be interested if anyone runs AIM using a similar approach, or any other observations that others might make with respect to such an approach.
Regards. Clive.
Hi Toofuzzy,
I created my own version of the Quick Calculator a year or so back using VB4. At the time I was asking Tom whether any Newport DOS versions were still available as I remained a DOS'ite up to that time.
I never really liked the nursing you had to do with Windows, preferring tools to work for me instead of having to constantly care for the tool. I'm not keen to run a racing car that needs constant maintenance and tweaking just to keep it running (or at least not crashing), instead I'm content to run a basic but reliable car.
I upgraded to NT4 back then, being on service pack 6+ and sufficiently stable, and have remained with that since. But I still drop back to DOS preferring the likes of Quatro Pro for spreadsheets and its flexibility with graphs.
Being an ex-programmer has installed an insight into just how unstable most of the newer stuff tends to be and as such I'm happy to let others put the effort in to iron out all of the problems before I'll move onto that platform.
With respect to AIM-LD, I intend to look into that more deeply over the next few days as if anything that is more along the lines under which I'd be using AIM. My current intent is to use the basic AIM, but assign a virtual amount to that, such that I might not actually have the cash as indicated in the AIM accounts, or possibly even some real debt.
Regards. Clive.
Re : AIM seems to act as the "trading" version of Buy and Hold or the "Buy and Hold" version of trading!
I'm now visualising AIM as more of a complete highly competent team. Comprised on an asset allocation manager who oversees, cash, buy and hold and range trading departments, collaborating to identify the best distribution of funds between each at any one time, helped of course by the current iWave value presented by Directorship. Collectively the cash department helps lower downside losses in a declining market, the trading team provide benefits during sideways ranging markets, and the buy and hold team help ensure any under-performance during Bull run's are acceptable, but maybe needing a bit of a hand with the Director stepping in with a VEALIE.
Never any in-house bickering or personnel problems, despite receiving no pay or bonuses, requiring little directing, and signalling actions in a mechanical like manner to best ensure a buy low sell high type overall stance rather than the buy high sell low outcome that often occurs when human emotions become involved in investment decisions.
Regards. Clive.
On extending the testing further Tom, the trading component of AIM does a pretty good job overall.
I ran some Bollinger band - 1 standard deviation ranging measures, combined with trend analysis so not to buy or sell against the trend, in a manner much like what a range trader might do, and there was a high correlation between the traders signals and that of AIM's when both BUY and SELL SAFE were set to 0 (so as to produce an above average number of trade signals).
My conclusion, for what its worth, is that there's little point in extending the complexity of AIM in an attempt to improve upon its range trading component as most likely any improvements seen would be a result of data mining - that is patterns in the test data, rather than any real value add.
I'd be tempted therefore to simply set Buy SAFE to 5%, and Sell SAFE to 10% as a initial default, so as to account for the 5% or so average capital appreciation that markets tend to rise by over time.
Regards. Clive.
Hi AIMster : Re : So what, exactly, is your methodology?
Buffett tells it all in section 7 of his owners manual. http://www.berkshirehathaway.com/ownman.pdf
His insurance business float is now around 70B against a 140B odd company value - in effect a 150% leveraged type overall position. With float (debt) costs of 0% then if his average investment return is 14% that equates to a 21% gain.
In my opinion yes he does do a good job of investing, but far too much discussion with respect to Buffett focuses on his stock selection methods and too little on his 0% leverage costs competitive edge benefit.
So my approach is to in part emulate that asset allocation strategy. Fully loaded to 150% exposure levels with periodic price driven stock/sector rotations, coupled with a long term view. However I appreciate that I can't secure 0% cost of debt and therefore carry a penalty in that respect. Providing however the return from my investing exceeds the cost of debt then a positive benefit arises. For example if I borrow half of my liquid value at say a 6% lending rate cost, but make 14% from investing the combined funds, then overall I'll see a 18% net benefit.
In my opinion spreading across asset classes simply serves to iron out short to mid term paper profit/loss volatility. Where you hold for the longer term and ignore such shorter term paper profit/losses its not worth paying that volatility reduction cost.
Complimenting equities with a modest element of leverage as and when relatively cheap levels of fixed dated debt can be secured, is sufficient to meet my needs. For example there are current opportunities around to secure debt at fixed 5% for ten years. I have a preference for above average yielding stocks and as such the dividend yield from my selections is alone in excess of 5%. Any capital gain in stock price on the leveraged component is therefore profit.
A difficulty with this approach however is that over time you have to secure more debt. That is as my total principle fund value increases over time, more debt has to be added to keep the 150% overall average leverage going. Otherwise in time you'll end up with perhaps just 110% or less overall exposure. Obtaining debt was a lot easier whilst I was employed than it is now. Despite having property and liquid assets vastly in excess of the level of debt I seek, being 'unemployed' deters lenders and consequently will likely increase my costs of debt over coming years (I will still likely be able to obtain loans, but will be charged a higher rate).
I've been aware of AIM for many many years and as such I'm now contemplating a revision to my strategy in view of my current 'unemployed' status to use a leveraged AIM approach. That is to assume a virtual 150% fund size, perhaps with 33% in cash and 66% market exposed say according to what AIM (or iWave) depicts, which equates to 100% of my actual fund value being equity exposed and a virtual cash fund of 50%. As such that would reduces my real debt requirements unless AIM starts flagging up to be more heavily exposed in equities when real debt would have to be added. Historically however I believe AIM rarely signals more than 80% odd exposure unless stock prices are extremely cheap.
Providing AIM achieves comparable returns to that of buy and hold, then there should be little difference in my overall average returns, whilst reducing the debt management side of the strategy.
So sorry guys, if, as would currently appear to be likely, I do start a migration towards AIM, then you guys will start seeing a lot more of me around here than my previous once a year of so visits ;>).
Regards. Clive.
As a quick 10 minute test, I downloaded from your AIM Software web page, Bill Riedeman's Excel spreadsheet for AIM. Using his already populated CAER0796 worksheet data for July 94 to July 96, over which the market price changed by a 24.7% average p.a. and the basic AIM with split safe set as per his values produced a 34.7% p.a. return on the 10,000 starting sum, running a relatively simple test of calculating the 4 week price based standard deviation for each review (row), and assigning an arbitrarily selected rule of stdev to safe setting levels of
STDEV SAFE
< 0.25 10
<0.5 8
<0.75 7
<1.0 6
otherwise 5
for both BUY and SELL SAFE levels, produced a total of 37.5% p.a. return.
In other words, as price volatility increased SAFE was reduced so as to narrow the hold range.
Comparing the number of buys and sells filled for the standard spreadsheet approach there were 10 buys and 7 sells, the variable split SAFE in contrast had 9 buys and 8 sells.
By far too small a sample to draw any conclusions I know, but perhaps indicative that a relatively simple modification can yield both bottom line results and better balance between buys and sells.
Regards. Clive.
Hi Tom,
I downloaded AIM Investor's trial version a little while back but it wouldn't correctly install for me. Kept prompting to enter a code which it showed, when I did, it would restart and again prompt for that same code again.
Do you still use Newport? As that link now appears totally dead.
I've asked this some time back, but don't believe you ever clarified Tom, as it was around the time that you were upgrading your PC, but what do you do with your spare time? You've been out of the rat race for far longer than me. I've been pretty well tied down with parental responsibilities over the last 3-4 years since having stopped work, and the couple of befriendments I've made with others who also have free daytime periods (also made possible through investing), is limited in that both are perpetual travellers - spending only a few months of each year in the UK before jetting off to Australia and Spain.
A hobby that you feel passionate about perhaps? Maybe AIM?
Regards. Clive.
Hi Toofuzzy
I've been aware of AIM for 20+ years. To me however it's a automated combined buy and hold, trading and asset allocation tool. Along the lines of having 33% in cash, 33% in long term stock and 33% moving in and out of the market, such that AIM has a tendency to smooth returns across the various market moods and in so doing tends to be seen as outperforming in all but a raging Bull type markets. As such I don't personally utilise it in practice as I consider myself to be a long term investor who is little concerned with any paper profits or losses evident over the short/mid term period - I simply don't need such short/mid term smoothing.
Neither do I trade in and out of stocks, however that said, for the trading based element of AIM, rather than using fixed levels of effective support and resistance lines at which stock is targeted to be added/withdrawn, some traders utilise the likes of Point and Figure's, Bollinger's or other standard deviation based models and I would have expected that such techniques might equally have been of partial benefit in the AIM's trading based area.
AIM-LD is I believe comparable to a leveraged based AIM. A potential benefit being that at a 50% initial cash:equity split, an effective 200% leveraged position would have 100% equity exposure from the offset, but carry no debt servicing costs - at least not until the equity proportion for the underline basic AIM indicator rose above 50% levels.
Before anyone takes exception to any of the above comments, be aware that no disrespect or critic is intended, each to their own. One of the best aspects of this board and Tom's endeavours over the many years, and the reason for my revisits, is the pleasant and informative nature that is all too apparent, yet so often all too lacking on other boards. Part of this arises I believe out of being able to consider alternative views in a mature manner over that of being a focal point of like-minded individuals who flame any non-conformance.
Regards. Clive.
Hi all. Re : SPLIT VARIABLE SAFE
Let's consider I want to use AIM on a Stock or Index. An Index typically ranges through around 25% each year - something like a 9500 low 12000 high perhaps. Individual Stocks however typically are more volatile and have perhaps a 40% yearly range. Knowing AIM likes volatility I might therefore opt to apply it to a Stock rather than an Index.
The next hurdle to overcome is what then to set the split SAFE buy/sell levels at. For simplicity let's assume a stock price has ranged through 100 low to 140 high. The mid value being 120.
Also assume that I wish to trade on a as-and-when basis, that is as price trigger levels are encountered, rather than on a fixed weekly/monthly/quarterly type basis.
Pragmatically therefore I take the 120 mid price and using the 100 - 140 previous years high/low range calculate that a Buy SAFE of 10%, Sell SAFE of 5% and a minimum order size of 10% would trap a AIM purchase signal at the historic years low and/or disposal at the historic years high.
Ideally I should perhaps uplift those ranges to account for the natural upward motion of stocks over time, but to keep things simple I'll discount that concept for the time being.
(Equally I could have opted to use a 5% Buy SAFE, 0% Sell SAFE and keep the 10% minimum trade size, which would have narrowed the trigger price levels down from 140 to 133 and lifted the 100 to 104.)
Sure enough over coming months the stock again cycles through a high and low range. It doesn't matter which way this occurs, it ends up with a 10% of current holdings having being bought and then later sold (or sold and later bought in the opposite direction) with a 40% separation.
I've achieved a 40% gain on 10% of my holdings as a result of such a cycle. If the cycle repeats several times over the course of the year so much the better, more combinations of 40% gains on 10%. That's a 4% benefit compared to the total equity exposed funds for each such cycle.
For each cycle of low to high transition, the round trip buy and sell costs might amount to say 5%, reducing the gain down to perhaps 35% on 10% of the exposed equity value. That's still a reasonable 3.5% pre-tax benefit.
Volatility (beta) in a stocks price is often as difficult to forward project as that of selecting outperforming stocks. Beta is based on a historical price measure, such that companies that have recently churned out unexpected good or bad news will typically have higher beta's to those that are ticking along as expected. Just because of a relatively recent piece of good or bad news doesn't however mean that high beta will continue into subsequent quarters, more often its other stocks that were ticking along with a low beta that suddenly go into higher beta. Usually beta is directional too. A large downswing in price or uplift in price to new price ranges levels and stays there until further unexpected re-valuations subsequently occur.
Of particular note however is that the Buy and Sell SAFE levels, together with the minimum order size are critical. They must be set at levels such that the troughs and peaks in prices result in actual in and out trades being made by AIM. If SAFE is too low, the gains are reduced. If SAFE is too high, then rotations are missed.
Setting SAFE levels according to historical price data would not appear to be the most appropriate approach. What then of a dynamic SAFE, adjusting according to the relevant shorter term oscillations in price levels. A 'split variable SAFE', that attempts to increase SAFE levels when too many trades are occurring such that tops and tails are being sold off/bought into too early, and reduce SAFE levels when there are too few trades occurring such that no cycle benefits are being achieved. Perhaps AIM might even be used itself to adjust SAFE levels according to measure of such fluctuations and oscillations.
I've seen Tom's AIM Algorithm Changes web page that details Split SAFE, where it is identified that so doing 'substantially outperforms' the by the book SAFE approach, but I'd be interested to hear if anyone else has considered or evaluated alternative methods of determining or adjusting SAFE levels as it would appear that there is great potential in such.
Fundamentally such a concept needn't involve be too much additional complexity. Minimum market order size is in effect a constant as we don't want to be trading too little amounts, leaving just the buy and sell SAFE levels as variables. A simple solution might therefore to have time decaying/rising based SAFE levels combined with an adjustment/fixing rule according to the number of buy and sell trades encountered, in an attempt to identify and localise around optimal levels.
Regards. Clive.
I know you internationally diversify more now Tom, so you may be interested to know that the UK's market progressed at around 16.5% total over the Jan 1985 to 2000 period - based on a Index fund's gross return comprised of a 12% average p.a. capital gain and 4.5% starting dividend yield level. If you'd been in a few years earlier, From 1980 odd, that figure would been more towards 20% gross levels.
In contrast the 1985 to Jan 2006 period is showing around 12% p.a. gross for LTBH, comprised of 7.5% p.a. capital gain and 4.5% start date based dividend.
In both cases I believe a fair estimate to be around half as much again of those total p.a. rates from alpha benefit.
However, I have year end price date for the UK going back to 1800 and dividend yield data from 1870 odd, having studied such over time I'm of the opinion that the 1980 to 2000 relatively strong gains arose out of the more widespread uptake of stocks and the upward revaluation of prices so as to be more reflective of what might be deemed fair price levels. Having achieved more realistic price levels however, going forward looks less likely to approach such 20 odd year historic average levels. Personally I'm anticipating going forward at around a 15%, alpha included, mid to longer term average rate compared to a 10% odd market average, in the belief that the last couple of decades 20% to 30% p.a. averages achieved with alpha are much less likely. With greater focus however, and hence my more recent interest in sector rotations, some further benefits might be achieved, as with so much interest and money around, I can't see the market wide volatility levels giving that much of a tailwind benefit.
Regards. Clive.
Hi Don.
It's not based on sector indexes but sectors. They don't actually clarify how the sectors were measured however I assume it was based on the individual sector averages referenced back over the years.
For 2005/2006 they opted to pick a couple of stocks in each of the four sectors and rotated in/out over the year in accordance with the following, achieving a 26% overall :-
First quarter Household (or Staples I believe they're called in the US).
Second Forestry
Third Utilities
Fourth quarter Electronics/Electrical
They've again picked a further two stocks per sector for the coming year in a similar manner.
I first read about the concept in the follow up 2006 article whilst in the library one day. Being sufficiently interested I subsequently obtained a copy of the first issue via goliath.ecnext.com - a charged report/article request service. I've since upgraded my PC and copied across only my own limited set of notes made from the two articles.
Regards. Clive.
In message 21035 you say 14% BEFORE income/taxes but in 21044 that changes to AFTER income/taxes. I'd be interested in a clarification as the former is about on par with a long term buy and hold comparison whilst the latter would be much more impressive, especially in view of a chunk of funds being in cash under AIM.
I ask because, as you know, I dropped out of the rat race around 3 to 4 years ago in order to spend more time with my two sons throughout their primary school years, and investment income is my only source of funding too. Despite having been familiar with AIM since the 1980's, I haven't actually used such indicators for many years now in the belief that the by the book AIM typically provides LTBH comparable returns but with lower risk, instead preferring to focus on asset allocation in order to alpha add.
A complexity for the UK a few years back was (and in part still is) the 0.5% purchase tax on stock, such that coupled with market makers spread and stockbroker fees, costs were relatively high for less liquid and/or volatile stocks, such that trading relatively small amounts as and when AIM flagged
was not viable. My personal preference is not to trade blocks of anything less than 10K UK (around 15K USD) so as to keep such costs relatively low.Over more recent years however such costs have fallen drastically, with tax free alternatives more readily available.
Many others over here tend to overlook such costs, one of the current popular choices being funds of funds, which are seeing inflows of tens of billions. In the case of hedge funds at the lowest level, their fees are 2/20, that's 2% p.a. and 20% of any profits made. Add to that the top level managers fees and ... well speaks for itself.
Regards. Clive.
Hi Tom. ADVFN is well known over here and one of the biggest financial information providers in Europe.
I wasn't aware of their buying out of Investors Hub and hadn't even noticed until you mentioned it. So as far as I'm concerned the access is little different to that of previous.
Their popularity may very well steer a greater number of participants from the UK towards Investors Hub, but personally I've had little dealings with ADVFN excepting grabbing the odd occasional graph. Their web site tends to be rather heavyily loaded with images and links, so it depends upon the exposure given to IH.
Primarily the UK yahoo finance web pages combined with the UK's version of The Motley Fool have cornered much of the wider market.
Together with the many financial spread-bet agents that tend to provide most of the required data/information feeds.
14% p.a.? Perhaps time to update the top Moderator (blue) sections reference to 20% p.a. Or has the LIFO 'by the book' faired better than yourself?
Regards.
Clive.
Hi Tom, Re Market Cycles
At the simplest level the concept tells use that sectors phase in/out in a rotational sequence of
Cyc, Tech, Ind, Basic Ind, Energy, Staples, Svc, Utils, Finance.
With Tech being the leader out of 'Full Recession' then the next in the set is Industrials during the 'Early Recovery' and Energy signifying the 'Market Top'..etc. As you say, AIM automates the actual identification of the actual phase.
With regard to diversification, I used to consider holding a single stock in each of a wide range of sectors, typically 15+ stocks, as being diversified, until someone pointed out the survivorship concept, that is historical price performance based measures tend to ignore failed stocks. My approach now is to hold a number of stocks in a much smaller range of sectors, around 4 or 5 stocks in each of 3 or 4 sectors. The principle being that if one stock fails in a sector, the remaining stocks in that sector tend to pick up on the failed stocks business (and potentially their assets at discount price levels).
As a simple example, say 3 stocks in a sector each priced at $10, making $1 profits - combined = $30 with $3 profit. One fails, the others pick up on the failed businesses trade and start making $1.50 profit and their stock price rises to $15. For the investor originally holding all three stocks they still have $30 worth of stock value and $3 profits being generated. As such the only cost of failure is a reduction down from 3 holdings to 2. If instead a single stock from a range of sectors is selected, then your more likely to hit upon unmatched failures and loss of value as a result.
Being more sector focused, as and when apparent cheap/expensive prices levels arise, sectors can be rotated in/out accordingly.
Did you see the 34% p.a. (20 year) average sector based strategy reference on the link in my previous message?
Regards. Clive.