fabijo's account talk

[fabijo]

I appreciate you putting time and careful thought into your spreadsheet, I tried digging into its inner workings but I'm clearly too dimwitted.

Maybe some time next week, or by the end of the month, I'll be done making so many changes. When I get the chance, I'll attempt a clear explanation and an in-depth explanation in the TA Tools subforum.

 

[fedgolfer]

before 9/03... VIX was calculated differently (see below). Not sure if this greatly alters the returns of your spreadsheet? The 2nd to last paragraph confirms how i've been using it the past couple months in the bull run.

from http://www.decisionpoint.com/TAcourse/AboutVIX.html


"In September 2003 the CBOE changed the way the VIX is calculated. The new VIX is derived from the prices of all near-term at-the-money SPX puts and calls and out-of-the-money puts and calls. Deep-in-the-money options are excluded. This methodology is also used to calculate the Nasdaq 100 (NDX) volatility index (symbol: VXN).

The old VIX is still avaliable but is now has the symbol VOX. It still uses the old calculation method, which uses the Black-Scholes pricing model. The VOX is calculated by taking the weighted average of the implied volatility of 8 OEX calls and puts with an average time to expiration of 30 days.

The VIX (and VOX and VXN) measures fear and optimism as manifested in options activity. When large numbers of traders become fearful, the VIX reading rises, and when complacency about the market reigns, the VIX reading falls. And since the vast majority of put/call buyers are wrong and lose money, it's usually a smart move to fade (go counter to) what the VIX says the the crowd is doing.

The VIX is an INVERSE indicator, which means that high readings are oversold (excess of bearishness) and low readings are overbought (excess of bullishness). Because of this Decision Point displays the VIX chart with an inverted scale to make its interpretation more intuitive -- overbought readings show at the top of the chart and vice versa.

Raw VIX numbers are of limited value. The VIX indicator is most useful when used in combination with some type of overlay, and preferably one that employs channels or bands. Some technicians use Bollinger Bands for this purpose, others use a short term (3 months or so) linear regression channel or percent bands. When used in this fashion, it is the VIX position within the channel that's important, rather than the raw number reading.

The CBOE calculated severals years history of closing values for the VIX. That data was not live and doesn't have an open, high, or low associated with it. Those days will display as dots on the daily chart. The same is true for the VXN, which is only a few years older than the new VIX.

More detailed information about the VIX is available at http://www.cboe.com.

 

[ayla]

before 9/03... VIX was calculated differently (see below).
from http://www.decisionpoint.com/TAcourse/AboutVIX.html

Raw VIX numbers are of limited value. The VIX indicator is most useful when used in combination with some type of overlay, and preferably one that employs channels or bands. Some technicians use Bollinger Bands for this purpose, others use a short term (3 months or so) linear regression channel or percent bands. When used in this fashion, it is the VIX position within the channel that's important, rather than the raw number reading. [/url].

Thanks for the very interesting info about the VIX. As I said in the other thread, I agree the VIX raw numbers are not enough. But if the 20, 50 and 200 VIX SMA values are added, from the charts I've seen, the VIX predicts "on the nose", the major and minor downturns that we have been experiencing periodically every three to five months or so (but not the smaller downturns that occur on daily or weekly basis or for other reasons such as news). Thanks again.

 

[fabijo]

Well, I am on my way to becoming a true mechanical investor. I'm going 100% G fund, because the numbers told me to do it. I will not try to figure out the market, I can only do what the momentum does.

 

[Foghorn]

Well, I am on my way to becoming a true mechanical investor. I'm going 100% G fund, because the numbers told me to do it. I will not try to figure out the market, I can only do what the momentum does.

Fabijo, you sound a little discouraged. Don't be. Remember, no one knows what's going to happen. You can only try to predict the market to smooth out the drastic drops, short term, or hold on to rollercoaster bar, for the long term.

If anyone could predict the market with 100% accuracy, they'd likely be doing something illegal.

I, for one, think you're on the right track. I've been reading your posts and they've helped me.

So, keep it up. You're doing well.

Foghorn

 

[fabijo]

Fabijo, you sound a little discouraged. Don't be. Remember, no one knows what's going to happen. You can only try to predict the market to smooth out the drastic drops, short term, or hold on to rollercoaster bar, for the long term.

True. True.

Thanks for the encouragement. I'm still not happy with my spreadsheet. I think I could simplify it more to do what I am thinking about. Once this week is passed, I'll be free from college courses until January. Maybe I'll be able to get the spreadsheet where I want it during that off time.

 

[fabijo]

Now it's time to hop on over to the C fund. This thing has me on my toes.

 

[fabijo]

I feel like I am chasing the funds around. That's okay for now. Today the sheet says to go to I fund. This thing jumps around almost every day. I'll be working on smoothing out the calculations so it doesn't jump so much. YTD return of this sheet (which I did not follow) is 22.3%. Let's see if I can get a better calc method that diversifies between funds and still generates a good return.

 

[ebbnflow]

Looks like your spreadsheet is on speed!

 

[fedgolfer]

Let's see if I can get a better calc method that diversifies between funds and still generates a good return.

Personally, i like your spreadsheet kicking out one fund. The obvious thing is that asset allocation will reduce your ROI and risk. But I like the clear green light it give for which of the stock funds is the most ripe for pickin'.

 

[fabijo]

Personally, i like your spreadsheet kicking out one fund. The obvious thing is that asset allocation will reduce your ROI and risk. But I like the clear green light it give for which of the stock funds is the most ripe for pickin'.

I would like to have it rank the funds. Then you could have the choice to go with the best fund or to split according to the ranks, like giving more percentage to the one with the highest rank. If I is WAY better than C, it could say 70% I and just 10% C or something like that.

 

[fabijo]

I'm going 100% S fund. I would rather go to the C Fund, because I think the bigger caps are going to do better. But, again, I'm sticking with the "system." By the time the weekend is over, I'll have updated my spreadsheet to use growth projections.

 

[fabijo]

Now going 100% C Fund on Monday. Now I feel a lot better about the spreadsheet. One of its weaknesses is that it follows the fund that has the steepest slope on the MACD Histogram. The problem is that some days the price will jump up real high only to correct itself over the next couple of days. That jump up would tell the spreadsheet to go to that fund. Another weakness is that on bottom days, the MACD is the most negative, so the sheet stays away from those funds.

Finally, I've worked a way for the spreadsheet to follow a bit of probabilities also. It mainly still follows the MACD, but I also added a sheet that calculates the projected growth ranges over the next two days. It compares today's price with the projected range from two days ago. If it breaks above the uppermost price by a certain percent, the spreadsheet either stays out of that fund, or gets out, because that is most likely a top. If it breaks below the projected lowest price by a certain percent, it jumps all in that fund, because the price has most likely hit some sort of bottom -- at least short term.

Now here are the results:

Prior to using probabilities:
June 2, 2003 to December 15, 2006
Monkey: 133.59%
Year to Date
Monkey: 22.9%

Now using probabilities along with MACD:
June 2, 2003 to December 15, 2006
Monkey: 157.64%
Year to Date
Monkey: 32.07%

Past performance is not a guarantee of future results, but I've got nothing else to go on. We're always referring to the past when we explain the now.

Here's a link to the new and improved spreadsheet. There are three pages. The one titled Real Prices is the one that will show you which fund to go into tomorrow. If you put in today's TSP fund prices, look at the allocation for the next day - that's the one to follow.

http://mircats.com/fabio/Projecting.Growth.xls
(Be aware that the file is about 10MB!! My server has been slow, probably because people keep downloading these things)

 

[fabijo]

Man, with the market making big moves then little moves, it just makes my spreadsheet jump up and jump around. I hate to keep doing IFTs, but I bet that tonight it will say to IFT to G or F.

 

[FUTURESTRADER]

Fabijo,

Do you think appyling the beta (or volatility) for each index proxy might be beneficial. I.e., (EFA) beta is 1.0 as is (SPY) (which makes sense since beta is the measure of volatility versus S&P 500), however (IWM) (Russell 2000 ETF) beta = 1.5, 1.5 times more volatile than S&P 500. So in your strategy, though the slope of the MACD histo may be the same for IWM and SPY at a given time, the % gain/loss would be 1.5 times greater for IWM.

 

[VirginiaBob]

Hopefully, the indicator will say G-fund, since the penny is likely to pay COB Wed. So if you get in before noon Tuesday, you should get it and then sell on wednesday.

 

[fabijo]

Fabijo,

Do you think appyling the beta (or volatility) for each index proxy might be beneficial. I.e., (EFA) beta is 1.0 as is (SPY) (which makes sense since beta is the measure of volatility versus S&P 500), however (IWM) (Russell 2000 ETF) beta = 1.5, 1.5 times more volatile than S&P 500. So in your strategy, though the slope of the MACD histo may be the same for IWM and SPY at a given time, the % gain/loss would be 1.5 times greater for IWM.

I haven't used the beta directly, but the spreadsheet does modify the MACD based on how fast the width of the Bollinger Bands is changing, which is similar to measuring its volatility. There's always more to do. I still haven't gotten to the point where the spreadsheet ranks each fund, then diversifies.

Really, the next part I am going to be working on is having the spreadsheet calculate trendlines and trading channels long term (and possibly short term). The closer a fund gets to the top of its trading channel, the more likely it is to go down - and vice versa.

 

[ebbnflow]

Any chance in incorporating Tom's Sentiment Survey analysis as part of the decision making process?

 

[fabijo]

Any chance in incorporating Tom's Sentiment Survey analysis as part of the decision making process?

:laugh:

It looks like a couple of people are posting the positions of the top traders, so that is a good indicator.

I could keep going and incorporate everything over time! I might get to the point of making my computer emotional. I'd have to start from scratch when my computer starts telling me its afraid or greedy.

 

[fabijo]

Hopefully, the indicator will say G-fund,

That's exactly what happened. Now it says go to G. It hurts to lose 1% in a day, but it's better than losing 5% to 10% for a month! Now I know to look at how close a fund is to the sheet's projected top. On Thursday and Friday of last week, all the funds were within 1% of the uppermost extremes of the projected prices.