| Index | Recent Threads | Who's Online | User List | Search |
|
|
 |
Portfolio123 » List all forums » Forum: Testimonials » Thread: Cancelled subscription |
|
Total posts in this thread: 31
|
[Request a Feature]
[Post new Thread] [Add To My Favorites] [Watch this Thread] |
| Author |
|
|
jpkernot
Advanced Member UNITED KINGDOM Joined: Oct 9, 2005 Posts: 223 Status: Offline |
After a couple of years and great expectations i am bowing out. There have been understandable losses from the recent crash in 2008, yet I am not conviced that many of the curve fitted returns that are seen here can be successfully duplicated. It appears that only two or three clever users are actually successful in making money with the rest - and I include myself - chasing our tails. There is a lot of 'fool's gold' in the database. I am not saying that the site is not useful - just that I have been unsuccessful in using it. Going back to active management which was frankly more fun and more productive. Wishing everybody well. ---------------------------------------- Thanks! |
||
|
|
theilman
Advanced Member
|
Sorry to see you go. Unfortunately, this is likely a function of too short a time span available to conduct our back-testing and robustness checks. Recent tools and work done by Marc on regime switching are helpful, but they do not get to the core of the problem. Adding ETF's, shorting capabilities, improved interfaces, etc., only became hot topics when the market turned down. What will happen is that these tools will be put into place, and many will experience poor real-time performance because the data history is too short. I'm willing to bet that by the time Marco gets true shorting functionality up and running, the market will have turned and everyone will forget about it (unless you really want to run long/short, which is a good idea from a risk-adjusted return perspective). I understand that there are daunting issues when it comes to extending the database, but it should be priority #1. It has been put off so many times that just introducing more "tools" to keep members happy won't make a difference to Marco's biz in the long run; members will leave. If it is too difficult to do it, Marco should just come out and say it. If it can be done, than all current resources should be directed to getting it done. P123 is a great resource, which is why I joined in the first place. But I have been around long enough to watch the discourse on this topic, and it's gotten to the point where Marco doesn't even respond to threads that mention database extension. To a certain extent, I can't blame him, but more bells and whistles is not the answer. More data is. Ted |
||
|
|
dcnelson
Advanced Member
|
It is not the data although having larger time frame would help. Quants using extensive databases far more robust than what we have here are getting the same results. Quantitative analysis for the time being has stopped working. At some point statistical relationships will take meaning againg. Just not right now. |
||
|
|
victorass
Advanced Member
|
Why is "p123 not working"? I hope nobody here is expecting to find a set of mechanical rules for a long only portfolio that generates POSITIVE returns in a market that has 98% of all NYSE stocks trading below their 50 day and 200 day moving averages? What I use p123 for is trying to generate models and portfolios that beat their benchmarks. It's up to all of us to try either capture that alpha outperformance by shorting the benchmark (or buying puts) or we don't do that and we live with the stock market beta. My personal opinion is most people don't like hedging because psychologically its very very hard to carry hedges when the market is rallying and one is underperforming. It cuts both ways; one can't have it both ways. None of this is p123's fault in my humble opinion. It's up to us to design portfolios that have a net long exposure that we're all comfortable with. I'm a happy camper, good trading all  |
||
|
|
charles123
Advanced Member
|
I think one of the major issues is people are not willing to learn how to incorporate outside tools into thier analysis. If you want to hedge with a short ETF. Dump the share price history from Yahoo into excel and run a model there with the price history downloaded for you trading system from the graphs page. You can do the same thing for a shorting system. You can combine and weight various systems, long and short in excel. Is it the exact same thing as having the capability at P123, no, but it;s good enough. Also, what possible good would it have done this year to have data from the 1990's incorporated into your analysis. If anything the data being generated right now during this market meltdown will be far more valuable in the future. You can't expect institutional level tools that cost $10,000 per month for $50-$100 per month from P123. The alternative is to go use Zacks backtester for $1600 per year with all its garbage data. Good luck getting a straight answer there. Just my $.02. |
||
|
|
theilman
Advanced Member
|
I understand what the last 2 posts are saying, but that has nothing to do with likely future success. I happen to work for a large endowment, and I can tell you that our quants that go long/short are barely down for the year. The discretionary managers, hedge funds or long only, are taking it in the shorts. I am not "blaming" P123 for shortfalls in strategy development. It's just that one can only do so much torturing a dataset of a certain length. Hedging is no more difficult to maintain discipline on than following a long-only strategy, if one has confidence that their strategy covers expected future outcomes. I for one am dying to see how Marc's work on regime switching holds up from say 1994 to 2001. Ted |
||
|
|
theilman
Advanced Member
|
1990's data would be quite valuable. Growth (large cap in particular) outperformed. How do you think most of the value-oriented strategies would have fared during the 1990's? Dodge & Cox, one of the all time great value managers, was getting fired left and right during this time frame because they didn't "get it." How about developing a regime swithching sim that incorporates not only risk premium and eps, but the ebb and flow between value and growth, and small and large cap? You can't capture that with the current database, whether going long or short. Of course one can download Yahoo data and short some ETF's, but it will be a purely technical system and one can do this with a lot of existing platforms for not very much money. More data gives more confidence. Beta risk will always be present, but I have seen where that can be handled either by a regime-switching type of application (one of our managers does this, but in a different way), or long/short. And I don't really get the $10,000 argument or the Zach's argument for being happy with what we have. Extending the database has been an active topic on this site for the last couple of years and no one was talking about relative pricing until recently. If Marco feels it necessary to increase the subscription 500 or 1,000 a year to cover while still providing the best value anywhere, I bet he would have a number of upgrades and more new subscribers. More data is a no-brainer. Ted |
||
|
|
jpkernot
Advanced Member UNITED KINGDOM Joined: Oct 9, 2005 Posts: 223 Status: Offline |
No, but once you have followed enough ports you will realise that only about 10% of the stuff here 'might' work better than your own judgement. I also find a lot of clamouring about 1-3 stock portfolios (not you Olikea) rather misses the point. Chase all you want but the fact is that even the best ports here have done abysmally over 2 years. Alpha was king but alpha is dead. ---------------------------------------- Thanks! |
|||
|
|
jpkernot
Advanced Member UNITED KINGDOM Joined: Oct 9, 2005 Posts: 223 Status: Offline |
Indeed that is also what i am seeing from pros who have the ability not to be shoe horned into a specfic startegy. Unfortunately P123 gives a false sense of security from thinking in a box. That box is simulations. ---------------------------------------- Thanks! |
|||
|
|
grokkalot
Advanced Member
|
IMO, the last 6 months has been a uniquely bad time for the type of approach to investing that P123 enables. Historical relationships haven't worked because "the market" has emphatically decided that things are different this time. Essentially all "risky" and/or "cyclical" assets have performed in a manner that is only paralleled by the early 1930s. Commodities and high yield bonds have done *worse* than they did in the early 1930s. Quantitative relationships based on trailing fundamental data or even rational profitability estimates don't work in an environment like that either. |
||
|
|
[Show Thread Printable Version] [Post new Thread] |
