Author Topic: Auto Invest for Folio  (Read 13458 times)

dontvote

  • Full Member
  • ***
  • Posts: 226
    • View Profile
    • Email
Re: Auto Invest for Folio
« Reply #15 on: January 24, 2014, 07:38:01 PM »
They are also free to buy loans from WebBank, which they in fact do for every single loan they issue.

*yawn*  No kidding.

Quote
LC is free to trade on folio for their own account as much as anyone else. Perhaps the word 'front-running' is too loaded - they are not doing anything illegal.


I'm not even talking about legality.  Screw legality.  I asked you once, I'll ask again:  What makes you think that Lending Club is front-running?  If they're front-running they're doing a piss-poor job of it when little ol' me can beat them once in a while.  Do you want to explain that?  Do you want to explain any of these comments of yours? 

Yet another person I have to start "a file" on.   lcdude you can come clean or we can all examine your life history right here in the thread.  You pick.

That's a weird and unnecessary threat Core.  There is no need to get personal. Save it for 4chan.
dontvote

Interest Radar IRR: 46.11%
NSR ROI: 41.22%
Average Loan Age: Your Moms

core

  • Hero Member
  • *****
  • Posts: 1790
  • Your loss is my gain
    • View Profile
Re: Auto Invest for Folio
« Reply #16 on: January 24, 2014, 07:57:00 PM »
Nobody got personal, dontvote.
He dropped some serious charges and I want to see some answers.  Rather than hit & run comments.

As for 4chan, I'm afraid I don't know what that is.

dontwatchkidsonline

Fred

  • Hero Member
  • *****
  • Posts: 1421
    • View Profile
Re: Auto Invest for Folio
« Reply #17 on: January 24, 2014, 09:48:35 PM »
LC runs their own folio bot that is frontrunning all other investors.

All investors?

I can show a proof that LC does not front run me.  Many times I was able to get loans when the FundedAmount = 0; which meant I was the first investor for the loan. 

It automatically snatches grossly mispriced loans before they even show up in the folio listings.

You seem to indicate that LC notes show up in the FOLIOfn listing automatically.  This is not true.  LC notes appear on FOLIOfn because some lenders put them for sale there.

Sorry, any attempts to implement an automated strategy for buying grossly mispriced loans will be fruitless.

Tried and failed?
« Last Edit: January 24, 2014, 09:53:21 PM by Fred »

core

  • Hero Member
  • *****
  • Posts: 1790
  • Your loss is my gain
    • View Profile
Re: Auto Invest for Folio
« Reply #18 on: January 24, 2014, 10:06:32 PM »
Quote
All investors?

I can show a proof that LC does not front run me.  Many times I was able to get loans when the FundedAmount = 0; which meant I was the first investor for the loan. 

Fred, this subtopic was about frontrunning Folio.  AmountFunded can _never_ be 0 by the time it's on Folio.  The other two points made sense. 

Quote
That's a weird and unnecessary threat Core.

Smile for the camera.


dontvote I've opened "a file" for you now.
« Last Edit: January 24, 2014, 10:19:12 PM by core »

core

  • Hero Member
  • *****
  • Posts: 1790
  • Your loss is my gain
    • View Profile
Re: Auto Invest for Folio
« Reply #19 on: January 24, 2014, 10:53:48 PM »
My, my, my dontvote.  You've been awfully unruly lately and now I know why.  Why would a grown man do that?  I guess he wouldn't. 

Bohb Daishi

  • Sr. Member
  • ****
  • Posts: 481
  • I eat free lunches
    • View Profile
Re: Auto Invest for Folio
« Reply #20 on: January 25, 2014, 01:29:21 AM »
I would reasonably expect that LC purchases a certain percentage of notes from the Folio platform. They, above all else, should know about the deals you can find there.
There are three ways to make a living in this business: be first, be smarter, or cheat.

dontvote

  • Full Member
  • ***
  • Posts: 226
    • View Profile
    • Email
Re: Auto Invest for Folio
« Reply #21 on: January 25, 2014, 09:50:44 AM »
My, my, my dontvote.  You've been awfully unruly lately and now I know why.  Why would a grown man do that?  I guess he wouldn't.

Quote
That's a weird and unnecessary threat Core.

Smile for the camera.


dontvote I've opened "a file" for you now.

should I be ready to hand over my lunch money to you before class tomorrow? please don't tell anyone that I cried after I tried to kiss you!
dontvote

Interest Radar IRR: 46.11%
NSR ROI: 41.22%
Average Loan Age: Your Moms

Fred

  • Hero Member
  • *****
  • Posts: 1421
    • View Profile
Re: Auto Invest for Folio
« Reply #22 on: January 25, 2014, 12:11:39 PM »
They, above all else, should know about the deals you can find there.

1. Notes availability in FOLIOfn is controlled by note-holders; LC has no way to know when a lender decides to put his notes for sale on FOLIOfn.

2. All the notes in FOLIOfn are available in one downloadable file (https://www.lendingclub.com/foliofn/notesRawData.action).  Anyone can write a program to download this file every few minutes to get a near real-time monitoring of the market.  With simple risk models, you can find good deals quickly & continuously.
« Last Edit: January 25, 2014, 12:13:40 PM by Fred »

Keltset

  • Full Member
  • ***
  • Posts: 176
    • View Profile
Re: Auto Invest for Folio
« Reply #23 on: February 02, 2014, 02:18:55 PM »
The problem with offering up automation software, which I have written for folio now that I use it more more than I had before, is the lost benefit to the author. If for example, myself or someone like core, were to release our python scripts with our filters etc then we quickly start competing against our own users. At what price would it be worth my selling my software? Well the price at which I start to make more selling the software than I do using it.

So what if I wanted the best of both worlds. Well, I can tell you one thing: Code I sold run intentionally run -slower- and be -less efficient- than any code I ran myself. Why would I intentionally want to reduce the odds of my obtaining those premium notes? Why would anyone?

The reality is that most of the people that run services are still backending the best for themselves and not using the same exact portal they are releasing to everyone else. And why wouldn't they?

If as one part of this thread suggested, cutting down your time was worth 100-300 a month for you. Well, the odds are that it would cost any of us running that software that much or more to release it (for every user that values that time saved).......

Sure there is benefit in releasing something when you are no longer able to properly capitalize on the available notes due to cash flow capacity. Then there becomes a value in selling access to sloppy seconds. Or maybe a programmer actually has no desire in investing in LC notes and is just looking to profit on the code.....

I have thought about releasing a configurable version of my folio python code but that would work on a build your own filter system, but again I find it hard to wish to compete with myself even throttling or charging per purchase for notes I myself haven't picked up first.

The reality is that folks like myself or core that have taken the time to write this stuff are doing it for their own benefits. I have spent a massive amount of time writing code first in one language then porting it to another as I needed to increase the speed that it ran at to properly compete to get the notes that are desired. Then there is also the other problem, I pound the hell out of the folio servers obtaining note listings.... It would be bad for us to see more and more people doing this as it quickly becomes a problem that folio would likely look at wanting to address or start throttling back users that are scraping the hell out of the site for the data.

It's a hard nut to crack, the desire to keep the code to yourself or to release it and make the necessary profit for releasing it.

Rob L

  • Hero Member
  • *****
  • Posts: 2137
    • View Profile
Re: Auto Invest for Folio
« Reply #24 on: February 02, 2014, 04:14:10 PM »
I have spent a massive amount of time writing code first in one language then porting it to another as I needed to increase the speed that it ran at to properly compete to get the notes that are desired.

Did you start with Python and move on, or start with something else and move to Python?
Python's very nice, but fast it isn't. Guess other options are even slower though.
If you feel the question is too revealing I apologize and encourage you not to respond. Just curious.

Your post does a good job of summing things up.

core

  • Hero Member
  • *****
  • Posts: 1790
  • Your loss is my gain
    • View Profile
Re: Auto Invest for Folio
« Reply #25 on: February 02, 2014, 04:51:00 PM »
I was wondering the same thing as Rob.  Out of the frying pan, into the fire.

The truth is though, the language itself isn't going to have a huge impact on speed when we are talking about these kinds of things with multiple network requests.  (Similar to how I always say it's silly to talk about buying servers close to LC's, shaving off 50ms when you are wasting entire seconds on things because of your flawed approach.)   What a particular language can do, though, is "encourage" you do to things in a manner that is slow.  It's not the language's fault, it is how it's used and in some cases the 3rd party libs that come to be associated with it.  Give someone a raw socket and some threads and they likely could get similar performance from any language.

Bohb Daishi

  • Sr. Member
  • ****
  • Posts: 481
  • I eat free lunches
    • View Profile
Re: Auto Invest for Folio
« Reply #26 on: February 03, 2014, 01:45:28 AM »
The problem with offering up automation software, which I have written for folio now that I use it more more than I had before, is the lost benefit to the author. If for example, myself or someone like core, were to release our python scripts with our filters etc then we quickly start competing against our own users. At what price would it be worth my selling my software? Well the price at which I start to make more selling the software than I do using it.

So what if I wanted the best of both worlds. Well, I can tell you one thing: Code I sold run intentionally run -slower- and be -less efficient- than any code I ran myself. Why would I intentionally want to reduce the odds of my obtaining those premium notes? Why would anyone?

That's the big question. Now, if you could sell your program to a bunch of people for a total of $5,000 a month, would you stop trading at that point? You certainly wouldn't benefit if it was just $500/month. But when it gets to the point of providing the same level of income as a full-time job for working just a few hours a week, is it worth it then?


edit: typo
« Last Edit: February 03, 2014, 08:38:59 PM by Bohb Daishi »
There are three ways to make a living in this business: be first, be smarter, or cheat.

Keltset

  • Full Member
  • ***
  • Posts: 176
    • View Profile
Re: Auto Invest for Folio
« Reply #27 on: February 03, 2014, 03:29:13 PM »
I started using a basic scripting language then moved over to python and find that its not really the language that causes speed issues its how you use it (as mentioned above)....

I use bits and pieces of libs but rip exactly what I need from them into my script. Primarily threading is the key as also mentioned above and when properly written and implemented I simply see no other value in going to another language. Personally, my choice to move to python was strictly having more experience with it and less fiddling around with learning to code via another method. Don't get me wrong, I'm not proclaiming python to be the end all be all of coding languages, it just works for me and does what I want it to with very high efficiency. Although I do believe Core summed it up very well, "Give someone a raw socket and some threads and they likely could get similar performance from any language." I take the approach of writing my own code and being a cut and paste coder which is the type of coder that often falls into the traps that many languages encourage and can help to create major inefficiencies in code.

To answer the question regarding selling the software to matching that of job, sure there is always a point at which the sale of the product could outweigh the personal benefit gained by keeping it proprietary. The question is at what point that becomes that valuable is still a simple, and yet very complicated, answer. Once the benefit becomes large enough to offset the direct gain of keeping it to myself (or any other authors self) AND covers the additional expense of maintaining and supporting code for external consumption which can also become much more time consuming than it is to maintain and support for yourself.

There are other issues to this as well. Selling software requires the investment of time into building sites, making the stuff user friendly to the non-author, and marking. The list goes on and on as to the effort it takes to go from self used and managed scripts that do what the author wants them to vs. taking those same programs public and releasing them as salable software.

There is something I always like to point out to people... Ever watch those late night television commercials or websites that say things like....

"Follow my program and you could be making thousands a day"
"These easy steps to automate cash flow"
"Simply follow my program and we guarantee you can make up to 30ka month..."

etc. etc. etc.

Then they are selling you these programs for 2,3,5 hundred bucks. Why? If these things worked so well why sell the "key"? Because they make their money from the buyer of these programs.

Most of the times things work very well -when kept private- but release the key and they stop working as well. Look at what information you have actually gained from the note flippers here. Sure, there is help and directions given, NJG released an article that had some information regarding one of his strategies.... Not a single highly successful note flipper here is going to release their secret sauce or their secret sauce becomes just another bottle of ketchup on the shelf and no longer produces the same benefit it does for them directly.

Sorry if it sounds a little blunt, but it's true. The "gravy" is always going to be kept to the ones that find it.

This doesn't mean there isn't value to selling the left overs, and in fact there may be many that enjoy and profit nicely from sloppy seconds. I can tell you that a 5k a month income stream would certainly encourage me to look at creating and releasing some highly configurable software that would assist users in implementing -their- own filters and implement -their- strategies. But it's also not my forte... I don't like sales, I'm not great at marketing, and to be honest theres many other things that I can do with my skillset to continue to increase my revenues.

There's one thing I have learned greatly from my experiences coding, it's not the primary market where the money is- it's the secondary market and that's where there has been little effort by others to capitalize on the inefficiencies of that market on any real scale. NJG wrote about how the other day he had 8 solid notes picked out on the secondary but by the time he checked his cart out he was only able to obtain 2 of those notes. If he had enough time to locate 8 notes- He should have owned all 8 of those. My personal guess? I know he uses IR and there are inefficiencies there that he simply can't keep up with. He should be running code that is optimized and written -for his strategies- to simply obtain those notes and simply obtained those notes the moment the code realized they were there. It takes no time at all to compare a note against a set of values to determine if it is something you want or wish to discard.

Rob L

  • Hero Member
  • *****
  • Posts: 2137
    • View Profile
Re: Auto Invest for Folio
« Reply #28 on: February 04, 2014, 04:34:40 AM »
Gee, so many things t agree with. While not fluent in Python I've used it enough to develop quite an appreciation. Don't know what schools teach as a first language these days, but a subset of Python would be a very good choice. From Python one can easily learn additional languages if / when the need arises. As an interpreted language Python crashes nicely, providing great detail as to the probable source of the problem. Of course as an interpreted language it is probably 10X slower than a complied language such as C for CPU intensive applications. The raw speed of computers these days for apps like those discussed here renders this moot. In the rare cases this is significant the Python/C API  provides the developer the opportunity to call C modules from Python where maximum speed is essential. I'm just an old C hacker having written  bazillion lines of C code. That's just what works for me, like Python works for Keltset, and something else works for others. Whatever maximizes your personal efficiency is probably best. A pet peeve is all those languages that protect the programmer against his/her self. ADA, Pascal, and so on started this trend. What I like most about C is that it is the complete opposite. Compilers have gotten much better over the years but C offers you almost zero protection against your own folly and it crashes ugly.

Personally if I were interested in selling software to P2P lenders (and for the record I am most definitely not) I'd avoid the execution side and go for some more powerful, flexible and graphical data mining and loan selection back testing tools. The tools that are out there now are a good start but just the tip of the iceberg of what's possible. Everyone is looking for an edge; something other folks don't know. Give them something cool to look for it.

What's out there now that I can try out Folio strategies and see how they would have worked over time? Given the nature of the Folio market and historic data availability is this even possible?

After listening to all the tax stuff lately there's probably a market for something that would simplify the process of keeping and reconciling tax records. It's availability would very probably increase the number of Folio users quite a bit. There's lots of potential Folio investors that won't use the service because of the huge tax mess every year.  I remember years in the past when I was very actively involved in the equity index options market (CBOE/OEX). Same deal as Folio with hundreds of trades, but at least the rules were clear.