Slope of Hope

I sometimes read this blog called ‘Slope of Hope’ by Tim Knight. Tim is a Silicon Valley guy who started a little business selling technical analysis charting software. Prophet charts or some such nonsense. I believe in TA but I think simple stuff works best, Support & Resistance, Relative Strength, Volume, etc. but none of the indicators/oscillators have any predictive value IMO. By far the most useless aspect of TA is ‘Elliot Wave Theory’, I won’t even bother explaining it if you’ve never heard of it, you’re better off that way.

In any event, Slope happens to be an extremely popular website. It caters to market Bears making it a great psychological attraction. There have been numerous studies in behavioral finance that individuals tend to gravitate to bearish scenarios much easier then bullish stories. One point I continue to harp upon is the importance in understanding human and behavioral tendencies. Prominently displayed on his site, even after writing the passage displayed below, is charts plotting the next ‘wave’ of some up/down cycle which is guaranteed to change the next day. The problem with EWT is that it has absolutely no statistical rigor applied to it and completely hindsight biased. Check out what Tim wrote:

As I sit here, watching gold and silver explode yet again to new highs, I can’t help but think of the tens of thousands of dollars in profits that I denied myself. And – - I really gotta say – - it just pounds a few more golden- and silver-plated nails into the coffin holding the body of my faith in Elliott Wave as a basis for forward-looking decisions. It’s all very cute and precious and lovely in retrospect, but for predictive value………..I’m really having grave doubts.

The posting is here. Guys, EWT is crap. There is something to say about cycles and patterns but EWT is utter crap….

Phi – Option Sensitivity to Dividend

I have a project to find a long term warrants sensitivity to changes in the dividend. It took me a good hour of searching to find that the name of the proper greek was Phi. After further searching I could not find anything else so I walked over to Borders (NYC is a convenient place!) and grabbed Espen Haug’s book. I finally found what I was looking for in Options Pricing Formulas:

Phi = -T*S*e^(b-r)T * N(d1)

b = Risk Free Rate – Dividend Yield

T = Time in Years

N is normal distribution

d1 is a big, long bunch of numbers I’m sure you can find anywhere

S = Strike price

Another Cool Visualization Technique

Intraday CL Trading Strategy

So the first few days of work have been interesting and somewhat stressing to say the least. It’s probably a good idea to write something about my experience in a future post. Since beginning my job I’ve transitioned very quickly from predominantly mean reversion to seriously examining trend-following systems. Primarily due to the reinforcement feedback on a trading desk with open collaboration. It’s much cooler to call out the beginning of a multi-point trend then scalp a megacap stock for .25 after going through 3x risk draw down. As such I did some work on crude oil futures and since I still trade equities I thought I would share the system here and hopefully get some feedback. My only request is that if you do utilize anything here that you report back how it is doing!! Seriously I’d like to know.

Posting Delay

I recently started a NEW POSITION and I’m still getting acquainted with everything so posting will be sparse but there’s a lot of exciting information and a number of different trading styles on this new desk.

Crazy Equity Curves Continued

Continuation of this post, I also found a toy model on the forum where you can play around with position sizing and portfolio allocation between cash, etc.

One consideration I forgot to mention but was also pointed out by an astute poster was that as capital grows in your “risky” strategy, you need a mechanism to rebalance capital. Otherwise there comes a point where the capital in ‘risk’ and the position size become a large portion of the overall portfolio capital and hence you risk even larger overall Draw Down. For example, if $50,000 of a $1MM account turns into say $100,000 and you risk 100% capital, well now your position size is ~9% of your overall portfolio…

What to do with crazy Equity Curves…?

Below is a quote I pulled of the TradeStation Forum from the same individual, Joe Joseph, that posted the webinar in the previous post.

Performing this exercise as intended with one year of live trading data and a hypothetical starting account size of $1,000,000 (for simple numbers) we get the following result:

It turns out that an account that starts with $32,000 and risks 31% on each trade makes almost twice as much as an account starting at $1,000,000 which risks 2% on each trade. An account starting at $75,000 risking 14% on each trade makes just as much as the $1,000,000 at 2% account. And this is just after one year.

Further, after two years, the smaller accounts start winning at the $250,000 at 4% level and the margin of victory grows dramatically for smaller, more aggressive accounts.

The specific numbers are not terribly relevant, but what is illustrated is that if you break your account into small chunks and trade each of those chunks more aggressively, you will actually make more money once enough time has passed.

What’s even more remarkable is that the risk to initial capital is actually lowered at the same time. As an example, the maximum possible drawdown on initial capital for the 4% case is 25%, whereas for the full $1,000,000 account traded at 2% the maximum possible drawdown is 100%.

Even better than this, a trader who chooses to risk the entire 1,000,000 account and is unfortunate enough to fail loses her ability to fund another system with this capital, whereas the trader who broke her $1m account into four $250,000 pieces and traded one of those pieces at a time with 4% has four opportunities to succeed.

There is more to be learned here, but this should provide a good start.

At first what seems to be a brilliant insight is simply mean-variance optimization. A financial adviser will suggest you make an allocation into a “risky” asset (stocks) as well as a “riskless” asset (CDs, Bonds). The net effect of this diversification is a larger absolute return over investing 100% in stocks or 100% in bonds.

However in strategy development, this methodology can offer some interesting alternatives. We no longer have to dismiss erratic equity curves offhand. Say we develop a strategy with fantastic absolute return but relatively large draw down; We can easily cut DD percentage in half by doubling the initial capital, right? Some might say that’s cheating! Well is it? Your primary goal at the end of some arbitrary period is to have more money then you started without losing a lot of sleep. Compare the various scenarios and find out if it is in fact true. Here’s one idea, place 90% of your capital in T-Bills and utilize the remaining 10% to buy/sell speculative options. Using this thought process we now have a use for some of this equity curves we would have dismissed previously.

Webinar

Below is a decent and short webinar (~45m) from John Jospeh of NextD systems who trades Trend-Following strategies on various equity indices.

https://admin.connectpro.acrobat.com/_a816688188/p40129421/

I found some interesting conclusions:

Sharpen Sharpe Ratios

In this previous post, I discussed some of criticism of sharpe ratios and particularly the focus on investors and prospective employers on a certain sharpe ratio. The problem with sharpe is it assumes ‘normal’ distribution of returns (return / std. dev of returns). This unfairly penalizes large percentage positive returns.

The good guys at the Yale School of Management wrote a paper and presentation on a method to manipulate or artificially boost the Sharpe Ratio of any strategy, useful for marketing purposes. It simply takes the existing portfolio and sells OTM Calls and Puts on the holdings or a corresponding index. This has the effect of cutting off the large positive portion of the return distribution and elongating the tail on the negative end:

If we break this down synthetically, the portfolio is actually long stock, short strangle. Further breaking it down, the long stock, short call is equivalent to a naked put, plus the additional naked put, the strategy is equal to short multiple naked puts. Amazing isn’t it?? Part of the presentation mentions a fund called Integral Asset Management which managed money for the Art Institute of Chicago and proceeded to lose ~$50MM dollars. The managers in question probably retained a nice 2% fee and 20% performance fee for losing a ton of money. How would you like to get paid to lose money? I would, I would! lol.

I also had experience with a recent short premium CTA, Zenith resources. The managers had an amazing fee arrangement, they charged 0% management fee and 30% performance fee. Brilliant! Your investors suspect the fund is confident enough in their ability to achieve a return that it forgoes a management fee. Meanwhile the manager actually just got their hands on a potentially very valuable call option. And the SEC is worried about insider trading, ha….

Experiments in Spread-Betting Part II

ME = IDIOT

This blog and its writings represent my public, although still ‘private’ identity, front to the world. As such I try to keep the writing as professional and error proof as possible. As many have likely gathered I do not do a very good job of that as it is so when I make an error in calculation or conclusion, it looks especially bad. A few days ago I posted an experiment I was conducting on exotic options and a new systematic strategy. Thank god I paper traded it because as I was reviewing the results today I had what a co-worker once dubbed as a ’senior moment’ and when entering barrier levels on a daily basis since Nov 3, I had an incorrect parameter as one of the inputs and hence, incorrect barrier levels.