Falling out of bed (quick update)

A quick update on yesterday's post. Yesterday the market closed below that support line. I spotted a new support line that had been broken seven days ago. That break brings us close to the signal level for a down call so close to the point where I start selling off on a 5% trailing stop and I start to build cash.

Nails bitten to the quick

The big question for me is: will that line hold? For the past three days it has teetered on the brink. Today is no different. The market has broken down but has not yet penetrated the lows of the last few days. Only time will tell. But I still have no negative signal.

Individual shares on the other hand are different. I have two that are creaking all of a sudden. GTN is really bad and EVC is almost as bad. I think I have to get rid of both. The question is do I wait till they recover a bit or do I ditch them now and take the whole of the horrible losses. I'll let you know when I decide.

Clear information is emerging about my stock picking systems as they work in the real world. The two top unisearches in my list (A and B) are indeed performing the best. A has generated the equivalent of over 50% return pa in the space of a month when the market has done nothing. System B is not far behind. System D on the other hand has generated some dismal returns.

The results are more or less as expected, but until you try in the real world you don't know for sure. It is almost time to take profits, rerun the systems and buy new shares. A good moment to ditch the shares bought with system D and go to systems A and B 100%. (See this post for the original work on this selection of unisearches.)

It's summertime and we have loads of visitors so posts are shorter and less frequent. Bear with me we shall soon be back up to speed.

How well have those share picking systems performed?

I have a plan

We’re now 9 days in from my first buys on the US market. The overall return has been 2.6%.

I hate the way that Vector Vest shows the Annual Rate of Return to indicate how fast a portfolio is growing over a short period of time. It gives you a clue about the rough trajectory of your stocks but there is nothing to say they will continue on that course. For what it’s worth and using VV's approach the portfolio is making a little over 70% p.a. Over the same period the VVC has fallen back by 0.3% or -9% on an annualized basis. The metrics on the performance of the strategy are good.

The strategy uses a number of different unisearches to pick stocks. See this post for the original analysis.

• The A search which performs best in backtests but is risky has done well. It has chosen 8 out of the 14 shares. 6 out of 14 are winners- average increase 3.5%
• The B search picked 3 shares : 2 winners 1 looser – average increase 6.7%
• The D search also picked 3 shares : 1 winner 2 losers – average loss 11.3%.

So what now?

I have decided to use my ISA account again to buy US shares. I still have some free cash and I will use the C search to pick my next lot of shares. Timing will depend on how the market is moving and my gut. 

We are in a consolidation period. My judgement is that it will burst out to the upside but I must be ready to be wrong and to know what to do if that happens.

My US system has two stop criteria for pulling out:

• The first is a four week stop. This means I will start selling shares on 15^th August
• The second is a 5% trailing stop which would be triggered if the market moves into a downtrend.

So I have a plan for the US market.

Those pesky UK shares

For the UK market I do not have a plan. The 25% stop loss and 50% take profit are my only strategies and those are based on guesswork and previous experience. The rest of what I do just has to be gut feel.

The two strategies I am using for picking shares are performing poorly (see this post for how they performed in back tests) particularly since they have been running for almost a month. The analysis does not offer much hope for an improvement over the next two months unless the market picks up strongly. 

The results are as follows:

• DBH has 2 winners and six losers average loss 5.8%
• 
  
• VSTPE has 2 winners 2 losers average gain 1.6%

A lot of thought is required about what to do next.

Hard work making money in the UK. Recipe: Insulin and Colcannon

Suddenly I remember how much easier it is to trade in the US than in the UK! My UK portfolio plods on sedately making me no money despite my best efforts.  That's not quite true. I have now completed the laborious backtesting exercise and have found  one, possibly two better unisearches.

But it's a week since I bought my shares and I'm still down 1.6% after costs. If I strip out costs and the ghastly THT I would be up 1.1% but those are unavoidable costs. I do not know whether, with a good stop loss strategy, I would have dumped THT before now. No Simulator, no proper backtestng, no decent strategy.

In the mean time in the US I got my buy signal on Thursday, I bought six shares that day and six more on Friday. (SOL, HSOL, III, UVE, GTN, EVC). Overall I have 7 winners and 5 losers after just 2 days. Profit after costs is 0.7%. There is no guarantee that this will continue but you must agree that it is a much better start.

I do have an idea for how to plug the gap left by the lack of the Simulator in the UK by concerted action by members of the user group. It would not solve the lack of a stop loss strategy but it would guide us to the best unisearches. Unfortunately I am not a natural organizer so someone else would have to drive the idea forward. Anyone interested in leading the exercise or participating should contact me.

Insulin and colcannon

I was first diagnosed with diabetes 15 years ago. That is what brought me into stock trading all that time ago. I retired with far too little money and had to make the most of what I had by driving my capital very hard. So far it has worked.

Since then I have collected a growing band of chronic autoimmune diseases. It seems that each time my blood is tested another one pops up. I am lucky in that, so far, none has significantly cramped my style, except in the range and number of pills that I pop every night and morning.

The progress of my diabetes is interesting. I started with dietary control which kept excessively high blood sugar at bay for  six or seven years. Then I started on pills. They worked for another eight or nine years. All of that time I had to be very careful about what I ate and, latterly, when I ate. The big risk then, as now, was falling blood sugar which carries the risk of a hypo which starts a little like drunkenness and can end in a coma. Being caught driving with a hypoglycemic attack loses you your licence so it's no joke. 

A few months ago my poor old pancreas, the organ that produces the insulin that manages blood glucose levels, finally gave up its efforts despite the help provided by pills. I was put onto an insulin injection regime. I tried two daily doses of mixed insulin which helps most people who can't be bothered with a more complicated system. With good advice I was steered onto an alternative system where I take one type of insulin at night and another with each meal. This keeps tighter control of blood glucose levels and gives me far more flexibility. But it is hard work. You have to calculate how much carbohydrate you are going to eat at each meal and adjust you insulin dose accordingly. On the plus side you can miss meals if you want to, and occasionally have big ones with lots of carbohydrates.

Which brings me to colcannon. I have always loved mashed potatoes, a big no, no when you are managing diabetes with diet. And colcannon is a fabulous Irish version of this dish. I was inspired by a visit to the Aveyron in France where we were given aligot, a local specialty where the mashed potato is mixed with cheese and garlic. It was a little bit bland for my taste but that might have been the versions we were served.

Net result I came home, took the appropriate dose of insulin  and made colcannon.

It's dead easy. You cook peeled potatoes so they are really soft. Don't ask me how many I used. I had a bag left over in the larder and I used what I had left. Cut up, I guess, there might have been a dozen small pieces. I also had some left over carrots which I steamed over the potatoes for about half of the cooking time. I had grown a cabbage, one of those pointy ones, and cut that up and steamed it along with some sliced industrial sized spring onions for the last couple of minutes. 

I then mashed the potato with lots of butter and salt (I confess to a liking for Aromat as a supplement to salt). It's a Knorr product favored on the Continent especially in Switzerland .)

I whizzed the cabbage and carrots in a blender for a few moments and then mixed them into the potatoes. I served it with sausages and gravy but it would go equally well with lots of other things, or by itself, or even cold. (I shall give you my gravy recipe one day.) Great plaudits.

What a difference a day makes?

The high DJI of May 22 has been broken. The move up is not super-convincing but it has happened. And for me the more important fact is that the market has given me my buy signal. My toes, or perhaps even my knees are in the water. I have bought 6 US shares They are KTOS LEE ARC ORBC SBCF and SNV. They come, three each, from my two top performing unisearches.

 My nerves are on edge because we are at a new all time high on both the DOW and the S&P.

It helps that my UK portfolio has improved. Five winners and five losers as of today. The ghastly THT gave me a bad scare and it has ended the day lower once more. It has broken through the 89 day moving average but it did recover some of its losses as the day went on. So I decided that fortune favours the brave and it got another day's grace. All this angst results from the fact that I do not have a UK stop loss plan. No Simulator, no backtests, no system.

So that's how things stand. I am 49% invested. 45% of the total is in equities and the remaining 3% is my short ETF in US treasury bonds. 61% of the invested amount is in UK stock and 39% is in US.

Decisions today, trading the FTSE again

I have reviewed some of my US trading. Although I still believe that this is best market in which to trade, US trading through my ISA attracts unacceptable costs. Put briefly, I am not allowed to hold foreign currency in my ISA (a tax shelter for investment which means that I pay no tax on capital gains). This means that I have to pay exchange rate costs on both buying and selling shares. This can reach 1-2% in each trade in each direction. It might be possible to justify this cost in a strong bull market but in a whippy market like this one I end up paying my broker too much to make a decent return. This means that from now on I will restrict my US trading to my SIPP (another tax shelter with different rules which allows me to hold foreign currency:  there are no exchange costs, and I enjoy the benefit of my returns).

This means that I have to switch back to trading in the UK. Unfortunately Vector Vest does not offer the same facilities to backtest strategies as they do in the US. I have no system which tells me when to enter or exit the market. I have to guess.

The FTSE broke an important resistance level on Monday.It continued upward and yesterday it passed my test criterion. If this move had occurred in the US it would have been a clear buy signal. So today I decide whether to buy in the UK. I've left my decision till late in the day. The market is pulling back, I feel bearish, the market is hovering on a resistance line, and the US market is showing weakness. I think I'll leave my decision till Monday.

The market is probably showing strength because the Bank of England announced that it is too early for interest rates to rise. This is interpreted as code for a continuation of the QE policy. A big deal for the market. It has certainly put the knife into the value of the £. I am thinking that in view of that news the market should have been even stronger. More information to make me negative.

Using Vector Vest UK to pick shares

But in the mean time I have been trawling some of the unisearches that I know work well in the UK. Even though I cannot test them with the same rigor that I can in the US tests give me enough information so I will know which searches to use when the time comes to buy.

Some are searches that I have developed myself, others are standard VV unisearches. I have tested them in good times and bad. To paraphrase Longfellow, in good times they are very very good, in bad times they are mostly horrid.

I have chosen time periods that include the good and the bad. The first table shows the return for the whole period and compares it to Vector Vest's own index the VVC/UK. The average shows how well these searches did overall in good times and in bad, weighted by the length of the time period. I have not fully identified the unisearches I have used but some of you can guess. They all do very well except when the market is in free fall as it was in three of my time periods (chosen deliberately to stress test the searches).

The second table shows the annual rate of return so performance in the different tests can be more easily compared. None of this shows how the picking systems would have performed if they had been subject to sensible stop loss criteria or had sensible entry and exit criteria. Hard to test without the simulator.

I think I know which unisearches I'm going to go with.

Robotic line drawing

Here we are back in no man's land. The market is failing to make a clear move. One day up the next down, and so on. It leaves me with very little to say.

I am, however experimenting with a new charting tool called ProRealTime. I think you can get the end of day version free but I am using a version provided by IG Index which uses the IG trade prices in real time. One of its features is that it draws horizontal and diagonal support and resistance lines automatically. Here is the daily chart for the DOW. You can see that it has drawn the same higher resistance line that I tentatively put on the chart on Friday. It's the highest red diagonal line. I've included my old line in blue but I'm going to ditch that as an idea. You see how the drawing of these lines is subjective and evolving. Their weakness and their strength all rolled into one.

As I write he DJI seems to be showing some strength, following through from yesterday. We shall have to wait and see if it holds.

More Vector Vest research

When the sun did not shine I used my time to do more research on Vector Vest. It has this fabulous tool called the Simulator which allows you to test theories about how best to play the stock market. You will have heard that the past is no guide to the future but that is rubbish. The past is our only guide to the future. It is why we revere experience. The wise are people who have seen seen it all before and have learnt from what they have seen. The past is an imperfect guide but it is the ONLY thing we've got. Used wisely it can be very helpful indeed. 

The Simulator allows you to test an idea over any period of time you like. It provides the best guide to the future if you look to see how your idea fared over a range of earlier time periods. 

The simulations include three key elements:

• a timing list which defines whether the market is in an up phase, a down phase or is neutral. I have used my S&R timing list (look back a few posts for an explanation) and one of VV's timing lists known as the GLB/RT Kicker
• unisearches which are VV's share picking systems. I have picked ones that I have found to work well or that VV favours. I have not named the ones I've used but have called them a,b,c etc. suffice it to say I have chosen them carefully
• stop loss criteria which have taken me out of my positions. I have used one rather unusual stop loss system which ensured I did not hold positions for more than 4 weeks before replacing them in up down and neutral periods. In down or neutral periods I also ran a 5% trailing stop but bought nothing and kept cash realized until the next up period. I found that these stop systems worked well.

I picked a range of periods over which to test these systems:

• 18 Sept 2007 to 11 Mar 2013 This took us through from just before the sub-prime mortgage crash to the point where the market recovered to the level of the previous high
• 28 Dec 2007 to 10 March 2009. The period of the crash
• 28 Dec 2011 to 11 Mar 2013 A recent 15 month period
• 28 Dec 2012 to 27 June 2013 The most recent six moth period

The table below shows the results of the simulation runs as annual average rates of return (i.e. simple rates not compound). At the bottom of the table I have shown how the DJI and the VVC, VV's own proprietory index performed over comparable periods.

The simulations show that;

• on average all the simulations outperformed the market  many times over
• even when the market crashed the S&R system was able to make a profit and if the exceptional performance of unisearch A is excluded the system broke even.
• The Kicker system halved the potential loss in that crash period
• The S&R system outperformed the Kicker system by 2.4 times and was winner in all time periods except the last six month period. If the exceptional performance of Unisearch A is excluded the positions would have been reversed 
• Unisearch A on average generated almost twice the return of the next best search on both timing systems. However, it was a high risk strategy, on some occasions it only generated one share and so offered no diversification. It was not a system on which to stake the whole of one's portfolio
• Returns were fabulous. 
• Even if we ignore search A on average the S&R system yielded 41% pa and lost no money in the crash. 
• The Kicker timing system yielded an average 15% when A is excluded, also creditable. More importantly it caught up in the last two test periods and was only about 7% behind when A is excluded

Conclusion: My early quick and dirty tests to see if the S&R timing system works are endorsed by this latest round of testing. The GLB/RT Kicker also works well. There are 4 Unisearch systems that seem to be solid performers over a range of time periods. This approach works. Further research is necessary to endorse these results and to refine them.

Wishful thinking

I'm glad I did that analysis of a bear market a couple of days ago (last post). I don't know about you but I found it an eye opening exercise.  It showed me how hard it is to interpret what is happening day to day without the benefit of hindsight and how tricky the market is. It does all it can to blindside you as it winds down and doing all it can to take your money on the way. I'm hoping that my new support and resistance (S&R) timing strategy will help my to avoid most of the pitfalls. The backtests which I ran to check out my theoretical  performance suggest it will be a good guide. I plan to do comparative studies of how well VV timing strategies compare with my S&R method through the 2008-9 bear market and I shall report back.

So where are we now? I pointed out the volume spike that occurred on the 21st and showed how often a volume spike warns of a change in direction. Rather disingenuously I wondered whether the spike was a late signal of the pull back that had begun a couple of days previously. That was just wishful thinking on my part. I've pulled out of the market and I was looking for confirmation that I had done the right thing. Instead the spike was warning that the bulls would return to take advantage of the pull back. On the Friday the 21st and on the Monday the smart money merchants were squeezing shares out of nervous bears and they have enjoyed a nice little run up since.

I just sit and wait. There needs to be a further move up before my signal fires. Am I missing a run up? Yes I am. But it could still be a false dawn.

I rarely remember my dreams but last night I dreamt of how I should set about reentering the market. I wonder what that means?

Vector Vest Seminars

Did you notice that my fulsome praise for David Paul's Vector Vest presentation is now being used to endorse his seminars. I am delighted that VV consider my candid comments worth repeating. David ought to have an easy job selling what is the very best market analysis product I have ever come across. But people have a hard time recognizing a good thing when they see it.

Anatomy of a bear market

I’m so glad that I sold out. It was a tough decision even though I had a clear signal. Since I sold my shares they have fallen in value by 3.4%, 0.8% less than the DJI. The fact that they have fallen less suggests that they were quality shares. There were 4 winners and 14 losers. This is an endorsement of my strategy: when the going gets tough head for the hills.

The question now is will I be able to get back into the market at a lower price and start to make money again. This is why I like a powerful decline and not one that meanders with little rallies. It’s those rallies that get under your skin. They are a notorious feature of bear markets.

Let’s look at the anatomy of the 2008-9 down-turn you will see what I mean.

1. Starting the 28 December 2007 we had a strong 12% drop to 22^nd January.
2. Then there were six weeks  of vacillation
3. Until a gentle rally kicked in and recovered about ¾ of the loss in just over two months, in fact the top value in that rally coincided with the 71.8% Fibonacci retracement. Spooky
4. Then another two months and a bit of decline pulled the market down to 80% of its high i.e 20% of its original value had gone.
5. Then another couple of months of indecision
6. The next two weeks took another 20% off the markets original value
7. This sharp fall brought out bottom fishers in force, they had an uncomfortable 7 days as the market went up but gyrated violently before all their profits vanished
8. The bottom fishers had another go for six days and then the market bit them back and came down to 45% of its start value
9. Then there were two and a half months when bulls and bears pushed and pulled, with the bears coming ot on top
10. And finally the last hurrah as the last 10% was taken off the market; and the bear market was over down 52% 15 months after it began.

I recount all this, not because I believe we are in for another bear market of the same magnitude, but to remind myself that when the market decides to pull back it is very difficult to navigate the twists and turns. Without the benefit of hindsight it is, oh so easy to be misled.

I've run a test using my S&R timing for that bear period. (Go back a few posts and you will see how this research works) Using my more risky share picking search the portfolio yielded 93% pa in that falling market. There were, however substantial risks associated with that rate of return . On more than one occasion the whole portfolio was invested in just one share, which happened to do extremely well. There was also a maximum drawdown of 24%. A big strain on the nerves.

The more conservative of my share picking searches did much worse losing 28% over the period. Still better than the 52% that the market gave up.

I know that some people think that because the economy is poised for recovery there is little chance of a new bear market. My thinking is different. The stellar growth we have seen, means that a lot of good economic news is priced in already. 

And then the market has to deal with two more negatives. The flow of easy money (QE) is about to come to an end. And interest rates are bound to rise and then investors will have an alternative home for their money.

Pessimistic? We shall see. 

Lacking objectivity

A weakness in the methodology I use to create my timing list is that it is subjective. It would be more scientific, more mechanical, more reproducible if the timing signals were based on algorithms that could be programmed into a computer so they were objective. Untouched by human brain. I would prefer a system like that, like the various timing systems offered by Vector Vest.

I assume that Vector Vest has pulled out all the stops to create buy and sell signals, up and down calls, whatever you care to call them which work as well as they can make them. It’s what those guys do best so I assume they have done all they they can. But like all systems that are based on historical data they have a tendency to lag and the calls they the make can be late. I've tested my S&R system against theirs over lengthy periods of time, periods that have been chosen at random to avoid the benefit of hindsight creeping in and my timing list works better. (I do not claim that the tests have been exhaustive so the results must remain provisional).

The S&R system is subjective but signals are given as close to price action as possible. With this caveat in mind it is useful to see how judgement is used in drawing those all-important lines.  Luckily the last couple of days have provided an excellent example.

In the chart you will see my original resistance line coloured in purple. It was drawn joining three highs, two before support was broken on 31^st May and one after. That resistance line was broken on 7^th June and became a support line on the 11^th. Price action following that break was not strong enough to trigger an up signal. Also the break came after just 5 trading days so the “ten day rule" (see earlier posts) would have been triggered.

Subsequent price action has made me  re-draw the resistance line basing its trajectory on a mixture of highs and market close points. At present price action suggests that the market is respecting that resistance. We shall know later today whether it holds. But if it fails it is still only 9 days since the break of support so the “ten day rule” still applies."Stay out of the market, sit on your hands, better safe than sorry" is still the message.

Attractions of the bond market

I can’t stop myself from looking for opportunities. You will recall that I pointed out how interest rates are rising as bond prices fall. "That must present a chance to make money," I say to myself. Have a look at this chart. 

It has risen 10% since the 1^st May. It is an ETF that shorts 20Year US Treasury Bonds. Should I jump aboard or have I missed the boat. As Oscar Wilde put it“ I can resist anything except temptation.” And here it is doing a nice little pull back from its high giving me a new chance to climb on. My lips are smacking. 

Better not. I’ll do a bit more research and perhaps have a go text week. Phew! that was close. Keep those hands still.

Why did I not think of this before when I first spotted the rising interest rate? "There’s none so blind as those that will not see".

Rice and peas

This Caribbean recipe is a huge favourite around this house. Dead easy to make. I've adapted it from one I found on the back of a tin of gungo peas, (a sort of bean from Jamaica). You can make it with red kidney

beans or any other bean that’s dark in colour.

Chop one onion and fry gently in oil. The original recipe calls for a deseeded chopped chili to be added  but my wife cannot tolerate chili so I skip that ingredient. It still tastes great. I then throw in 300 gms of basmati rice and let it fry with the onions for a bit. You can add salt and pepper at this stage.  Just keep it on a moderate heat in the same pan. I then add a can of coconut milk and about a pint of vegetable stock. I then allow the rice to absorb the liquid and add water if necessary if the rice needs more before it is fully cooked. As the rice gets close to being cooked I add a tin of gungo peas or any other type of bean and the same quantity of frozen peas. The frozen peas bring down the temperature so the whole lot needs to kept on the heat till it is warmed through.

It’s ready in about 15 minutes and it’s delicious

A little lesson

I sit and twiddle my thumbs. It does not come easily to me and it does not make for very exciting blog posts. I could just type in "nothing to report" over and over. Instead I thought I would  explain a bit about Fibonacci and Support and Resistance (S&R). I don't really hold with Fibonacci but S&R is at the core of the new timing system which is going to guide my entry and exit from the market from now on. Rather usefully the chart of the Dow Jones has provided excellent examples of both systems over the past few days.

Fibonacci 

I'm pinching a lot of this stuff from Wikipedia but I summarize. If you want more just follow this link and this one.

Fibonacci lived between 1170 and 1250 so it was a while before stock markets came into being. He was a

mathematician. He recognized how much easier it was to work with the Arabic numerals than it was with Roman ones. He was lucky not to have attracted the attention of the catholic church for if he had, I have no doubt he would have come to a sticky end. As it was he advocated the use of the numbers 0-9 and the extension of that sequence by the use of place values. He demonstrated how useful this was in bookkeeping and other types of accountancy. 

He also popularized a sequence of numbers that had first been described by Asian mathematicians and he gave that sequence his name. The sequence runs 1,2,3,5,8,13,21,34 and so on. You've already guessed how it works, each number is the sum of the previous two numbers. Fibonacci used it to predict population growth in an idealized rabbit colony.

What is amazing about the sequence is how often it appears in nature. The branches on trees, the leaves on a stem, petals on flowers, segments of a pine cone regularly develop in numbers that are found in the Fibonacci sequence. 

Another interesting characteristic is that the ratio between each pair of sequential numbers in the list converges to a ration known as the golden mean: approximately 61.8%. This ratio is widely used by artists and architects to create works of art and buildings exhibiting outstandingly fine proportions. The ratios between other numbers in the sequence also converge to  fixed values.

Stock market watchers, at a loss as to how to predict changes in direction in prices in the market, have adopted Fibonacci ratios to make their guesses as to when the market will turn.

On the chart of the DOW you will see that I used a Fibonacci sequence to predict, reasonably accurately the turn in the market from up to down. I drew a line between the latest high and the latest low. Helpfully my charting program calculated the Fibonacci ratios between the top and bottom of that price movement and, low and behold, when it got to that magic 61.8% the market stopped rising and began to fall. 

I have to confess that I'm not a true believer, but sometimes it works. And when you've got nothing else to help you, you might as well give it a go.

S&R

On the other hand I do believe in support and resistance. There are programs which generate S&R lines but mostly they pick up horizontal lines.which are based on previous highs and lows. Like the one I drew at 14869 which acted as a stop for that last fall in the market.

A series of diagonal lines, linking a combination of daily lows or closes marked the support of the market as it rose to its high point. When it then broke down a new support line, which started at the new low held the market up for five more days until it was broken on 31st May. By then a new resistance line had formed which kept the market on a downward path until it was broken just five days later. And now that resistance line has become a support line and has helped the market in its efforts to move higher again.

This development of S&R shows how we can interpret market movements from day to day. The S&R lines guide our thinking very effectively. As the price movement approaches a support or resistance line we can infer that a change of direction is likely as price action bounces off the line and returns to direction of the prevailing trend. If instead price action moves on through the line we can infer that the trend has changed.

At present we appear to have entered a period of market indecision with trends lasting just a few days at a time. By contrast the last uptrend continued, pretty much unbroken, for over 100 days. It was kept in play by a whole series of support lines. There were short periods when the trend hiccoughed, either because there was a small pullback, too unimportant to count as a downturn, or because the trend accelerated and a steeper support line needed to be drawn. 

The point is that these support and resistance lines can be drawn as they develop and they are valuable as guides to entering and exiting the market.

At present the message is stay out of the market and wait for a clear up signal before making any new investments.

Short change

Times like these caused me to give up writing my blog last time. I sit and I watch with no money in the game. No profits, no losses. I have to remind myself that the reason I've pulled out of the market is: no losses.

I have conducted a few new backtests and, reluctantly, I've convinced myself that I am doing the right thing. I have taken my S&R timing list and tested what would happen if I bought contra ETFs (that is Exchange Traded Funds that are designed to go up when the market goes down) whenever there was a down signal and vise versa. 

I've tried this every which way, with stop losses , without stop losses, with restrictions on repurchases when positions get stopped out and without. I have tried using Vector Vest’s contra ETF watchlist to inject variety into the selection, and I've tried plain vanilla with a watchlist that just contains DXD and SDX (which mirror the DOW and the S&P). And I have tried a couple of Vector Vest’s own timing lists.

The answer is always the same. You can’t make money on the short side of the market. That’s not quite true. If you know in advance that you are in for a really big fall then you can make money, quite a lot of money, but how do you know in advance that the fall is going to be big?  You could do it by instinct but that’s plain gambling.

The results bear out the outcome of my own clumsy efforts to make money on the short side of the market. I’ve ALWAYS lost.

I’m not saying it can’t be done. I’m saying I can’t do it and now I’ve shown myself why it’s so hard. Best sit on the side-lines and wait till the market promises you an easy ride by going up.

Fibonacci power

It looks as though the S&R approach is calling the market correctly. Broken support has presaged a decline. It is very evident on the FTSE.

The Dow looks a little different. Yesterday the 61.8% Fibonacci held it down. (Sometimes those horoscopes are spot on). And today futures are indicating a 120 point fall to 15112 That would place it on the broken resistance line. It would be ironic if that resistance turned into support. We shall soon see.

As far as I’m concerned I’m hoping for a continuation of the fall. The bigger the fall, the bigger the bounce. A big bounce is what I need to recover the losses I made when I misguidedly, tried to short the market and to buy some gold as a hedge against renewed inflation.

Non Farm Payroll

Today should be an interesting and key day on the US market. At 1:30pm London time, (8:30am Eastern) the US Government will publish its employment figures(aka Non-Farm Payroll). They have been trailed for most of the week. On Monday the Purchasing Managers’ Index gave a positive signal, suggesting that the market was in better health than had been thought. On Wednesday ADP, an independent research outfit, released figures that showed an improvement in employment though a smaller improvement than had been expected, yesterday unemployment claims were down, in line with expectations. Today the actual employment number will be released.

Predicting how the market will react is tricky. In theory a good number, indicating improving economic health, should be good for the stock market. In these perverse times it will probably have the opposite effect. A poor number will be hearten buyers. They will interpret it to meant that the Quantitative Easing program will continue for longer. The supply of cheap money will be unabated and they can continue to buy. The expectation is for a very modest change, but the ADP number would suggest that the number could be below expectations. Result: we could see the market spike upwards. Not long to wait and then we’ll know what the numbers tell us and how the market will react.

Support and Resistance (S&R)

The core of my new timing system for buying shares is S&R and the last few days have given good examples of how well S&R works.  It is the latest break of support that has provided me with a signal to exit the market. A new horizontal support line (which I drew a couple of days ago) has brought the pull back in the market to a halt. Time will tell whether this is temporary or longer term. If it proves to be significant, i.e. the market resumes its rally my ten day rule will come into play and I will have to wait until a the market has been above a new support line for ten clear days before I can start buying again. This will keep me out of the market during any whippy period that develops. (See 2^nd June post under the heading “But in the mean time”)

More backtesting

I haven’t screwed up the courage to go back and create a longer timing list so I can backtest further but I have broken down the results achieved over the past 5 years by Picker b) (see that same 2^nd June post under “The big bucks come from share picking” and “Stop press”) to provide a year by year result. The years run from March 24 to March 23 each year. (Date picked at random.) The figures come out as follows:

	S&R timing list/Picker b)	DJ Industrial Average
2008/9	100%	-38%
2009/10	167%	41%
2010/11	90%	13%
2011/12	21%	8%
2012/13	33%	12%

Remember that this is a very aggressive picking system. Often finds few or no shares that match its criteria, shares so there are periods when there is little spread of risk. Shares are never held for more than 4 weeks and most are in very small companies. On the other hand the timing system and its failure to find shares  has kept us out of the market for most of the ghastly year 2008/9. This is not a system for the faint hearted but the results are extraordinary.

New Label

During the next couple of days I shall create a new label. “Vector Vest Research” will quickly bring together all the posts which report on the development of my Vector Vest trading system. You will be able to find it in the label cloud to the right of the posts.

So where are we now?

I'm not the only one to struggle in calling the market's next move. All the commentaries I have seen have little to say. Few say, as I do, I don't know. 

I have been persuaded to put a trend line on the graph, with 2 standard deviation confidence lines. I'm not sure they are particularly helpful so I have made them rather faint. What they show on the FTSE chart is that the last high of 5 days ago touched the higher confidence line and this was followed by a pull back to the trend line. There the market sits. It could go as low as 6500 without breaking the up trend. If you were to look at the DOW or the S&P you get very much the same picture.

The Nikkei has already penetrated the bottom of its channel and is looking rather sickly to my mind.

A few days ago I drew attention to the rise in the interest rate on 10 year US Treasury bonds. That rise continues. It is now touching the top of its trend channel. 

You will recall that low interest rates have been one of the drivers of the stock market’s relentless rise. But the chart shows that they have risen by 32% in the past three and a half weeks. Before the rise, the yield on the DJI shares was 1.7 times that on those 10 year bonds but it has now fallen to 1.3 times. The chart reminds us that bonds have yielded far more the past. The low interest rate motor for higher stock prices could be running out of steam. On the other hand since the interest rate is reaching the top of its trend channel , at least in the short term, there may be some respite.

What I am doing

I’m still holding the better performing shares from the portfolio I built up earlier this month. For those who are interested they are LGF, KR, SMI, PPC, BONT, NRG, CIG, AOL, NVTL, SIRI, FHCO, all on the American market. (Details available on Yahoo Finance )They represent about 30% of my portfolio with the rest being in cash and GVC, my stalwart UK share which is not doing much at present. A couple of these shares are probably for the chop later today.

Backtesting.

You will recall that I have been backtesting market timing methods and that I created a Support and Resistance system for timing entry to and exit from the market. One of the reasons that I am holding on to my remaining shares at this inscrutable time is that the system's sell trigger has not been activated.

The system's backtest yielded great results but suffered an unacceptable maximum pullback in when the market whipped back and forth in 2011. I have been trying to refine the system to reduce the amount of profit the system sucked back as it generated too many signals.

So far I have failed to find a method to reduce the pullback significantly. But the stop losses that I added to the system have increased the annual return from 36 to 45%.

The chart shows the pattern of returns generated by the modified system over the period November 2008 and May 2013.

Best laid plans

I lost my nerve. It was that fall on the Japanese market that really got to me. A 7% drop in one day. It just shows you what can happen when an overheated market hits the buffers. I was there when the tech market bubble burst and I will never forget what true panic felt like. I had pulled out before the real trouble started but I saw friends desperate as they struggled to get hold of their brokers. That candle big red candle reminded me of those frantic times

There was follow through in the rise in interest rates on US Treasury bond that I mentioned in the last posting too.

The Wall Street, spooked by Bernanke’s comments that QE was to be “tapered”, did not react nearly so badly but it was enough to frighten me. Everyone knows the adage “Don’t play with more than you can afford to lose.” Well I do that all the time. All the money I have left in the world, apart from my investment in my home, is exposed to the stock market. So when I find myself staring into a pit I have to act. I have a big appetite for risk but only when the odds are in my favour. And that’s why I had to abandon my beautifully laid plans to wait for a fall of x% after the break of support before I exited the market. I could not afford the loss it would entail.

One reader described how he allowed tightened trailing stops to take individual shares out of the market leaving him with very decent profits. Well done! I followed a similar path selling off my weaker shares and the ones which were losing money fastest on the day following that big market drop. My pain was in the loss of paper profits. I pulled out over two days reducing my investment from 75% to 33%. This was a more measured exit from the market than the one in mid-April. And that is partly because I now know that a break of support +x% is a good signal for exit, and that point had not been reached.

A friend asked me why I used the DJI as my signal rather than the S&P. There is no good answer to that question. I just happened to choose the DOW. It is the index that everyone talks about. What is important is that it worked when backtested. It is a signal just like any one of those favoured by technical traders. It’s like “buy or sell when there is a moving average crossover” or “when the MACD reaches a certain point do this” or “when price action creates a flag chart pattern consider this action.” What my S&R signal says is that when support or resistance on the DJI is broken plus a certain percentage, that is a sell or a buy signal. It has the virtue, compared with the technical analysis signals, that backtests show that over a long period of time S&R works exceptionally well. You will make money.

So here I sit, partially invested and with a decent pile of cash, waiting to see what the market will do next

. We are in uncharted territory. Without a map we just have to feel our way forward. It is only Elliott wave aficionados who dare to predict where the market will turn. I am not a believer in their approach which seems to me to spend much of its effort on self-justification as the market sails past their predicted turning points.

I have plenty to do. I need to extend my backtests of S&R and I need to refine the system so it is not so vulnerable to periods of loss through drawdown. That should keep me busy. 

At last the sun is shining and there is blossom on the apple tree outside my window. Long may it last.

Good news!

In fact it's fantastic news. I am very excited. Very, very excited.

Some years ago, I bought, at huge expense, from Vector Vest, the most fabulous tool. It's worth every penny. It's called the "simulator". So far I have used it to choose the best share picking methods out of the almost 300 that Vector Vest provides. It has guided me well. Unfortunately it only works for US stocks. However I have now used it to test timing strategies and in this area the US is usually a good guide to what is happening.

I spent yesterday tediously putting the simulator through its paces to discover whether my support and resistance (S&R) timing system works (see the previous post), and wow! Does it work.

The simulator allows you to test timing lists against each other. Each timing list consists of a series of dates which are assigned the value up, down or neutral on that day. You can then set up a simulation which instructs the program to buy or sell shares based on one of the search method provided by the program. The simulation starts with a $100,000 virtual portfolio.

There are several pre-loaded lists based on VV’s timing systems, e.g. “confirmed calls”, “primary wave”, “green light buy/relative timing kicker.”. To this I added a list based on my S&R  plus or minus x%. The chart below shows an example of how this works from 2010 when the market was choppy and buys and sells were close together.

I told the simulator to buy shares, based on my favorite search method , whenever the signal was up and to sell them and go into cash when it was down. The search tool I used was one I had picked out as being strong and consistent through a series of simulator tests over a period of years. The simulator would then wait for the next up signal and would then run the search again and buy shares from the new recommended batch.

The results I achieved were very heartening and, subject to further tests, suggests I might well have found that holy grail of which I wrote in the previous post.

I tested from 24 November 2008 to the present (4 years and five months of rising but choppy trading). The Dow rose 79% over the period. The returns achieved by my simulations were as follows:

	gain for period	annualised gain
S&R	164%	37%pa
Primary wave	-14%	-3%pa
Confirmed calls	77%	17%pa

Cracking result for my S&R you have to admit. I tested primary wave specifically because one of you kind readers suggested it might be better for my style of trading than confirmed calls. You may be right but the system seems to be hopeless for this type of long term trading. Thanks anyway. Please keep those comments coming. I need all the help I can get.

Here is a chart of how the simulation worked out for the S & R system. You can see that there were lots of worrying pull backs in profits, i.e. lots of times when in the real world, as an investor you would be sweating and kicking yourself, wishing you had pulled out before. The worst was a $53,000 pull back or 27% drawdown, as Vector Vest would put it. But the nerves of steel are a prerequisite for this game. 

A final caveat. Drawing and interpreting support and resistance lines is subjective, an art.  Vector Vest’s systems are mechanical. So all their errors can be blamed on the machine.

Lots more work is needed to check the results and to refine the system. But so far it looks very good indeed.

Next week may be an even thinner week for posts than this one. Lots of family stuff coming up. But I will endeavor to comment on anything dramatic. Here is last night’s chart. Talk to you soon guys.