Home > Tools of the Trade > Educational Resources > Fundseeder - my experience

Fundseeder - my experience

  1. I decided to give fundseeder a go. In case you missed the earlier posts this is a platform where you can have your returns and trades sent, a bit like collective2. Because the returns come directly from your broker other people can have more confidence that they're "real", although of course you can still run multiple accounts and drop the worst.

    Unlike the latter however this isn't a 'follower / social trading' site where you can get paid by people for copying your trades, instead the goal is that you might get picked up by potential investors or featured in Jack Schwagers next book. FS even intend to offer an incubation structure for qualifying traders.

    Since I'm not looking for investors my main interest was in getting access to some interesting analytics, seeing how I am doing against other traders with the same style (systematic / futures / technical), and who knows maybe making it into the book.

    As well as linking your account to IB so fundseeder can see your future trades and returns, you can have your returns history uploaded by IB, right from when your account opened. You can also upload unverified data, but it will be treated as such.


    First the bad things:

    Linking the account was very straightforward. But getting the historic returns transferred was a complete and utter nightmare. This required numerous emails to both IB and fundseeder to actually make the transfer happen. Apparently initially they forgot to send part of my position records. The problem is that nobody at fundseeder seems to do anything unless chased - they aren't proactive about getting back to you when there is an issue and getting it fixed. I found the IB guy very helpful in contrast.

    Even once transferred I found that there were certain days where I had massive returns (both up and down) that were wrong. I quickly realised the problem is that the returns calculation only adjusts for a movement in NAV caused by changes in cash, but not for position transfers (if they haven't fixed this, then my top tip is that you can improve your FS track record by gradually transferring positions into your account from elsewhere...). There is a nice feature to request return corrections, which I duly used, but again it took a lot of chasing on my part to fix.

    To give you some stats I registered and requested linking / uploading on the 24th February.... and finally got the correct returns uploaded today, the 10th March.

    Historic trades aren't backfilled just returns. So trade statistics are only filled in from the point when you link the account. But given the horror I've had with historic returns, this isn't a bad thing.

    AFAIK only other fundseeder members can see your returns, which could be other traders or presumably these mysterious potential investors. This kind of limits it for my purposes, since it would be nice to post a link so that people on ET could see my returns in near real time.

    The documentation is a bit lacking; I mean the site is mostly easy to use and there is a pdf describing the statistics they calculate, but apart from that you're on your own.


    The good things:

    Unlike collective2 you only need to provide administrator access to your IB account. Because c2 assumes you might want to copy / follow other traders, rather than just lead, it wants your full trading privileges. Listen buddy, nobody gets full trading access to my account even if they promise not to use it without permission....

    The analytics I can see are pretty good, with a nice range of benchmarks - and more to come. More below. As a naturally competitive person the leaderboard feature is very pleasant. Again, more below.


    A quick tour: Leaderboards

    The home page hosts the main leaderboards, plus what I assume are aggregated account curves for different kinds of strategy and asset class. There is also a 'leaderboards' page, where we can look at a leaderboard both overall, plus for any subtype of strategy / asset class. Incidentally there are around 170 active accounts on the site. The first 5,000 will get free registration so there seems to be plenty of space yet. The subtype strategy / asset class is self declared when you register so might not be accurate / consistent.

    It's possible to click through to individual accounts and get pretty comprehensive stats. See "my account" below for a detailed view of what these are like.

    [​IMG]

    The person icon shows you where you are in the various leaderboards. The default ranking is by 'FS' score; which as I understand it is a sharpe ratio adjusted for statistical significance; thus people with more consistent returns over longer periods will be more highly ranked. I rather like this measure, and I've used something similar myself. And this isn't just because I'm currently ranked second overall... (and #1 in technical and futures trading)

    You can also rank by other metrics. Out of interest the best sharpe was 3.73 with an FS score of 58 (the same guy had the best sortino and annualised return). Here's the account:

    [​IMG]

    This account had quite a recent bad drawdown (50%) from what looks like a classic negative skew strategy. The risk control also leaves something to be desired.

    In contrast the highest FS score of 87 had a sharpe of 'only' 1.42 but over 4 years and with a much more consistent performance and very consistent risk:

    [​IMG]

    My own inconsistent risk targeting up to January 2015 (see below for details) may be penalising me a little, but so be it.


    A quick tour: Fundseeder analytics

    You can get even more interesting stats for your own account using the 'analytics' function.

    [​IMG]
    Here for example is the one month rolling return for my account (blue) plus the newedge CTA index (in red).

    Same thing, for rolling six month sharpe ratio.

    [​IMG]
    The period is configurable. You can also plot sortino, volatility and daily GPR. There are numerous benchmarks available.

    There's also a drawdown plot. I'm less keen on the "equity curve analysis", as I don't think trading the equity curve is a worthwhile activity (see my blog). Interestingly the plots prove my point:

    [​IMG]

    The faint red line is below the solid blue line; this is bad for equity curve trading. You can change the parameters are overfit to your hearts content. There are several methods for doing this; this one uses something called Bollinger bands which I thought was a kind of champagne

    Many more analytics are in the pipeline. I love this kind of stuff, but could never justify the time to write it myself, certainly not with such a nice interface.



    A quick tour: My account

    This allows you to see your own statistics, plus edit your details, and of course as I said above report incorrect data!

    Here are the kinds of statistics you can see here. These are also available for any other trader, by clicking through from the leaderboard:

    Cumulative return
    [​IMG]
    [​IMG]
    Underwater chart
    [​IMG]

    Annualised vol
    [​IMG]

    You may notice that the drawdown, vol and return figures are lower than the ones I report on my own thread. That's because you have a choice of using the actual account value to work out % returns, or a fixed notional value. The latter has to be greater than the former. Since at times my account has been much larger than the amount of capital I have at risk (started at £300K, now £400K), I've used a fairly large fixed value of £1.5 million as my notional capital. This also brings my volatility down to more typical hedge fund levels (if anything a little lower than most CTA's).

    As you'll know if you followed my thread for the first few months I only had stock in my account and wasn't trading futures. Then I started trading futures with an annualised vol target of £150K (50% on £300K); this subsequently increased as I ramped up my capital at risk with profits. In January 2015 this dropped to a more reasonable £100K fixed maximum vol target. The changes since then reflect strength / weakness of forecasts.

    Account size history

    This shows the actual value of your account. Any withdrawals / deposits will be reflected here, but not in your returns.


    Performance measures

    [​IMG]

    Summary

    This is a nice idea, some parts of which are well executed, other parts still have their teething troubles (in particular the historic data upload). Having made the initial investment of time to get this working I'm looking forward to using this in the future.

    GAT
     
  2. Thanks for taking time to post.
     
  3. Very cool, lots of good information to aid one on seeing what can be improved. I use a Bollinger Band on my long term trading, so when losses get down to lower band, I increase size a little.

    Am guessing most of c2 members have very small accounts, cause anyone with decent account not going to allow anyone access. I really don't understand why they would need full access?
     
  4. Yeah, the top guy on C2 appears to be buying and selling single S&P 500 contracts. In contrast about half the people on FS have more than $100K. Seven have more than $1 million. One guy has more than $12 bucks (maybe this is Jack himself?), and is running 8th overall on the leaderboard.

    I did ask C2 about the access. Basically they have one standard "API"* method to link to interactive brokers, which they use regardless of whether you're a leader or a trader (copying). Even if you don't want to autocopy someone elses system (which requires C2 to trade your account) you still have to sign up to this.

    * I don't know whether it's really an API, or just a legal agreement of some sort.

    Of course you can use C2 without linking your IB account. But then you have to set up a seperate method to "trade" your C2 account. This means learning another API, or entering your trades manually on their web system, or using their strategy language.

    The amount of information you expose on C2 is also much larger - literally trade by trade. Personally I'm okay with that, but many people wouldn't be. The way that they rank and analyse results also seems less professional than FS.

    GAT
     
  5. Thanks for the detailed report. I am a FundSeeder user, too.
    Question: it looks like you are rated #2 in the leadersboard. Has FS approached you with the offer to manage seed capital yet?
     
  6. No. Probably for two reasons -

    - I only got my data uploaded yesterday, literally an hour before I wrote this post.
    - I ticked the 'I am not interested in outside investment box'

    Also possibly as I'm UK based; I suspect that the FS incubator is only good for people who need US regulation (I'd need to be regulated in the UK).

    GAT
     
  7. Ok, thanks. Well, keep us updated on the developments. You have an excellent trading record. Is your trading systematic/automated, or discretionary?

    Unlike you, my participation in FS is mostly about incubation.
     
  8. Oh, I've just looked at your profile. Are you Robert Carver?
     
  9. Yes

    GAT
     
  10. Why would you not be interested in outside investment, with such a good track-record ? Has your strategy reached the upper limit of money it can handle without adverse effects on performance ?
     
  11. No. This is the same strategy used by large CTA's; medium speed trend following with average holding period of a few weeks. Capacity is probably around the 50 billion mark. I'm not quite at that level.

    If I had outside investment then I'd on the upside (a) possibly earn more money; and on the downside (b) have to put up with a lot of crap.

    The problem is I don't need to earn enough money, I'm lucky enough to have plenty, and I'm happy with what I already have.

    Having worked in the hedge fund industry, I'm fully aware of the stress involved with running other peoples money. I deliberately stopped doing that in 2013; I don't see why I'd want to go back to it, for a significantly lower income than what I used to earn (at least until my fund reached a certain scale, assuming it ever did - most small funds fail). In fact it's worse than that, because I'd have to deal with all the regulatory and other nonsense that comes with running your own fund, which as a PM in a large fund you are protected from (I think the regulations in the US, for offshore CTA's in particular, are much less onerous).

    Of course there are some advantages to running your own fund rather than working in a large one; which I don't need to enumerate.

    I think it's fair to say that I'm a bit different from the average person on this forum who is probably trying to make the journey in the other direction (retail punter to professional manager).

    Mind you I would never say never. I would for example consider taking a deal where I was given the best of both worlds: as a portfolio manager in a large fund that worried about all the non trading admin and where I got a reasonable basic salary; but where I was allowed to run my own office remotely how I pleased. I did have quite advanced discussions with a large fund about doing this.

    However they correctly recognised that I would be much more valuable working within the fund's office. I don't consider myself to have an edge or something nobody else can do. What I did was to similar to what they were already doing so there was no point creating a new sub fund. What I am good at (I think) is making a lot of small improvements (or removing errors to be more precise) to the basic CTA strategy. This is something I'm better off doing within a portfolio management team.

    Alternatively I might be a good fit for someone running a multi strategy setup who wants to get into the trend following CTA space and add that particular string to their bow. I've had some approaches in that respect, but the guys running the fund weren't a good personal fit.

    GAT
     
  12. Cool. I've designed and developed my own ATS from scratch (I am a software engineer and a trader), and have read quite a few books on systematic trading. I look forward to reading your book.
     
  13. Man, that's a long post....I kind of hate reading.
    So -- what's the bottom line...you make 17.80% annually :D
     
  14. Just look at the pictures then

    GAT
     
  15. Your sarcastic reply is true -- that's what trading is; it's not so much about numbers and data....but rather just staring at pictures/graphs...and trying to decipher the meaning behind it. o_O:confused:
     
  16. I am a professional trader trading my own account, I love the idea of FundSeeder and was briefly a member, but a few issues led me to cancel my account.

    The first is the privacy of trade data. Their User Agreement and Privacy Policy do not properly guarantee that FundSeeder will not sell, publish or distribute trade data in ways I do not like.

    There is also no guarantee that trade data will be held in encrypted form and no clarification about which employees can access it and for what reasons.

    There is therefore a risk that trade data will at some point end up in the wrong hands and be misused, especially if you truly have a unique, working strategy.

    The execution here seems to be lacking otherwise as well. For several months nothing happened on the site. They did not send even a single email to their members. It appeared there was some personnel turnover and all development work on the site had stopped.

    My attempts to get in touch with their customer service and CEO to discuss the issues I had were met with complete silence.

    In protest I removed their rights to access my account. Unless there is evidence that people are actually getting funded via the site, I think I'll be quite slow to return.
     
  17. FundSeeder is the result of the acquision of what used to be RapaCapIntro, of which I was a member. At that time, the service was well managed, and I got to know the CEO and the site developers/quants through email correspondence and some coloboration. Now that the ownership has changed, the quality of service (in particular member service) appears to have gone way down. My email inquiries about the improper import of data are yet to be addressed.
     
  18. Sorry I didn't mean to be sarcastic - the pictures alone give you a good idea of what the site is like, hence my comment.

    GAT
     
  19. As far as trade data goes I don't think this is a big issue.

    http://www.elitetrader.com/et/index...opinions-useful-not-risk.294748/#post-4186893

    I agree that the service level isn't great, although part of me feels I should acknowledge that we're getting this for free, so maybe expectations shouldn't be too high. The CEO has read this thread (he's contacted me on the back of it), so he'll be aware of your concerns.

    GAT
     

  20. They got back to me, so we are talking. I find it a little unsettling that they still have the elementary things wrong. For example, here is the cumulative returns chart for my account:

    [​IMG]


    Clearly, this chart wants you to believe that I made 254 million percent gain in 3 days! Yeah, if I am so good, how come they don't give me OPM?

    Now, let's look at the "account size history" chart. My account does not have enough history, so let's use your account (hope you don't mind). Here is the chart (the black circles are mine for emphasis):

    [​IMG]

    So, is your current account worth $2,126,450 or is it about $600K as the chart shows? Not clear at all, is it?
     
  21. Hi NL

    The discrepancy between NAS and assets over time in my account is not an error.

    If you read the FS information you can choose to eithier have a fixed NAS (essentially the asset base value used to calculate % returns from the actual $ returns you make) or use the actual account value. The actual amount in my account (the lower chart) is (a) varying and (b) always much more than I actually have at risk (between 300K and 400K depending on the period involved).

    So for example the withdrawal of capital I made in November 2014 would have resulted in my % returns
    going up and my risk profile changing. But in reality the amount of money I'm deploying would have stayed the same. So the fixed NAS method made more sense.

    A fixed NAS has to be higher than the account value, hence my choice of a relatively large value.

    So the answer to your question is I currently have about £600K in the account, but the NAS is over £2 million (the reason for the odd number is to do with a currency translation as you can only specify a NAS in $).

    For your returns it's most likely your NAS is too small. So even a very low $ return is an amazing %. I had a similar problem in that my NAS went in initially with the wrong scale (other direction - my % returns were very tiny), and yes I had to complain to get it fixed, and yes it was a pain in the ass.

    I agree it's frustrating, and these kind of issues could be picked up with better quality control and basic communication ("Hey you put in your NAS as $2 billion / $2. Did you mean to do that?"). This takes time but is quicker in the long run than having to repeatedly fix problems because they weren't right first time.

    For everyone's information I had a long chat with the CEO of fundseeder today. Of course I can't guarantee they'll get this working overnight (like most startups they're probably trying to do too much, and they're probably under resourced), but I can tell you that they're definitely aware of what they need to fix, and the fact they were so keen to talk to me indicates they trying to do this.

    I'm even more convinced that these guys have exactly the right approach to this problem and I'd really like them to succeed.

    GAT

    PS Just in case anyone is suspicous I'd like to put on the record that I have no financial arrangement with these guys; this is my honest opinion. I'm not a shill for FS*, just a customer who likes the idea but is fully aware that the implementation is far from being spot on.

    * OK this is exactly what a shill would say. But shills, in my experience, tend not to write long reviews including a certain amount of criticism; short effusive posts are more their style.
     
  22. I've opted to use "Net Liquidating Value" instead of "Constant NAS", so my 254M% return still does not make any sense.

    Your account size of $2,126,450 does not make sense, either. As you've identified, it's £600K. Even when adjusted for currency, it's still wrong. I understand from your explanation that NAS is set arbitrarily by you, and is not reflective of the actual account size. Perhaps FS should come up with a better name for it, i.e. call it "Scaled account base", or something like that. That should be next to the figure that they should call "Actual account size".
     
  23. Yes maybe the name is confusing. But the account size isn't wrong; it's exactly where I set it. This facility to set a different account value is actually very important and useful. Again an example of a good product, but not very well communicated.

    GAT
     
  24. Yes, I understand the concept of setting the "constant net asset value", but it's still not clear to me whether the calculated returns are adjusted for it, or not.

    Let's take your case. Suppose that as of Dec 31, 2015, your actual account balance was £600K. FS says that your monthly return for January 2016 was 5.11%. If your calculation method was "NLV", it would mean that you made £30,660 in January of 2016. Now, let's say that you chose to use "constant NAS" instead, and set it to £1.2M, i.e. twice your actual balance. The question is, would FundSeeder report your January gain as 5.11%, or perhaps as 5.11/2 = 2.555%?
     
  25. It would be £30660 / £1.2m = 2.55%

    GAT
     
  26. I just joined fundseeder. Great potential, but I'm not seeing much in terms of analytics besides measuring the total value of your account over time. I would have expected some trade-level analytics, or at least info. like "average number of positions held overnight," "largest exposure (as pct. of account value) to a single underlying", "average number of securities traded per day," "commissions paid," "total monthly turnover," etc.

    Do they really not have any portfolio info. (trades, sizes and allocations) tracked, or have I just been unable to find it thus far?
     

  27. No it's not there. This is something that keeps popping in and out of Beta testing. When I first registered they had trade level data then it vanished; then I recently logged on via my smartphone and there was trade level data there (which was incorrect incidentally!) and then it vanished again.

    This is a free platform, and they do seem to be focusing more on selling funds to outside investors at the moment; to be fair that's the only way they're going to monetise the free analytics.

    GAT
     
  28. Yeah, although it seems that if you were a prospective invetsor, you'd want to be able to see trade-level stats and portfolio characteristics before even deciding if it was worth your time to investigate further. So I am kind of surprised they are not prioritizing it.
     
  29. May I ask you question: Did you sign up there as a trader or as an investor?
     
  30. I'm a trader. If you go to the leaderboard, that's me at the top :) [for today at least!]

    GAT
     
  31. Great thread and info on Fundseeder. Thank you for sharing your insight @globalarbtrader.
     
  32. Globalarbtrader,
    First thanks for the info you've provided on this thread, it's very useful. I know that you don't want to manage OPM, however I am curious if you've been approached by anything at or through Fundseeder to do so? One of their big premises is that they've got this nebulous bunch of funders who are going to "discover" genius traders and fund them. If someone with your record over this amount of time (and who was a successful professional trader, so kind of an undercover test even though you didn't mean it that way) isn't at least contacted by a funder, I'd say that entire premise of Fundseeder isn't exactly true?
     
  33. Yes they have approached me. But I said no.

    GAT
     
  34. Thanks!
     
  35. That means you are a regulated trader, right?
     
  36. Nope, not any more (I used to be in my old job). That was one of the reasons I said no to OPM. I have no interest in the hassle of being regulated.

    GAT
     
  37. Rofl
     
  38. Anyone else skipped the first post ?
     
  39. I thought it was a great post. The ability to take the time and focus on one thing for more than five minutes is a lost trait these days.

    Thanks for starting this, GAT!
     
  40. ...and a necessary attribute for successful trading.
     
  41. Does the track record on fundseeder starts from the date the IB account is linked, or from the date of the inception of the IB account?
    From your post, it appears that it's from the inception of the IB account, correct?
    Thanks.
     
  42. Yes, inception of the IB account.

    GAT
     
  43. Is this the default way?
    If I only want to track the performance from the day I link my IB, while ignoring the previous performance, is it possible?
    Thanks.

     
  44. No idea. You'll have to ask FS as there is certainly no menu option.
    But...I doubt it. Would make the system very easy to game.

    GAT
     
  45. They give you two options: since inception or since link date. This is reasonable I think as people may have changed strategy or restarted trading after a gap or such like. It looks as if the FS score penalises shorter performance histories to compensate for this. For example, GAT's performance gives him a FS score of ~80 but there are accounts with better statistics (including +ve skew accounts) but shorter histories with much lower FS scores.
     
  46. If you can't do this at fundseeder then you could consider to open a so-called "linked account" at IB. This will get a new Uxxxx number. Then connect your fundseeder to this newly created account.
     
  47. Can FS see the positions or only the performance history in the account?

    Thanks.
     
  48. Is anyone aware of any fundseeder seeded funds?
     
  49. Interesting silence to this important question.
     
  50. It does seem plausible there are not many yet. If you do a google news search on fundseeder it comes across as a startup just beginning to sort itself out (appointment of new CIO, CSO and external funding just recently for example). I also received an email from them recently which suggested things are moving forward.

    So, it's too early to read anything into it I think, basically.
     
  51. So, FS are pushing a "Virtual Summit" (free) for traders to attend online (https://www.summit.fundseeder.com/welcome). Surely if you are good enough to be trading real money profitably and serious enough to be interested in funding that list of speakers is a bit, you know, "retail"?
     
  52. FundSeeder needs to revise their performance metrics by which they rank the trading accounts. In their leader-board, the #1 ranked account has a 1.7% annualized return, 3.5% cumulative return over the last 2 years, and the $4,800 account balance.

    That's especially ironic, given that the FundSeeder rank calculation formula is "proprietary".
     
  53. Well it seems to be more or less a version of the probabilistic Sharpe Ratio (ie it doesn't just look at average returns over standard deviation of returns but incorporates skew and kurtosis of those returns). This makes sense and helps to penalise highly negatively skewed, high kurtosis strategies like selling vol.

    But yes they should also consider, if not already, penalising high correlation with the s&p500 and high correlation with a simple compounding bank account like structure. That would weed out that #1 account and also penalise many of the leaders which are in effect just long stocks in massive bull market.
     
  54. Yep, my account jumped from #33 to #16 in one day last Friday, and all I had was a very small gain on that day. So, it does appear that many of the top FS accounts have a high correlation with the S&P 500 index.
     
  55. I agree with penalizing the negative skewness, but the penalty for the high kurtosis is debatable.

    The FundSeeder's definition of kurtosis is wrong. Here is what it says:
    Kurtosis is a measure of peakedness (or flatness).

    And here is what Wikipedia says:
    Kurtosis measures outliers only; it measures nothing about the "peak".

    Now that we got this out of the way, consider a return distribution with a normal left tail, and a fat right tail. This would have the high kurtosis (by Wikipedia's definition). Would you really want to punish a trader who has a large number of abnormal gains?
     
  56. That would give a strongly positive skew which would increase your FS score. If you look at the definition of sample kurtosis on the wikipedia page the powers used are 4 and 2. So kurtosis is a measure of the "symmetric" fatness around the mean (ie the value of kurtosis increases for negative and positive deviations from the mean more than that measured by variance/standard deviation).

    Does the gain in FS score from positive skew balance out any effect of increased kurtosis? Not sure, but there is "a" balancing effect.