[BTC-TC] ET.DIFF: Future to speculate on next difficulty change

[eltopo]

Great, it will be nice to have more future-like securities to speculate with

I did notice a minor typo in your contract:

0.01 + (current difficulty - last difficulty) / last difficulty) / 100 * 2

You have 1 opening bracket and 2 closing brackets.

Thanks, it's corrected now.

[Deprived]

Security seems fine - though only responsive in the range -50% to +50% per change.

The upper end isn't an issue but the lower end is.

I don't think anyone expects difficulty to FALL in the coming months, so allowing for it to in the payouts is inefficient.

The formula for payout on LONG is :

0.01 + ((current difficulty - last difficulty) / last difficulty) / 100 * 2

That means a payment of 0.01 if difficulty doesn't change at all - and more if it rises.  That seems wasteful in an environment where a fall just isn't likely - it means half of all capital invested is paying fees without ever being exposed to any useful risk.  In effect the shares are half the price and an additional extra amount is being tied up, shuffled around and having fees paid on it when everyone knows it'll just go back to LONG at the end.

I think if you remove the "0.01 +" and change the 2 to a 4 you'd have an instrument which sold at the same price but where the range for LONG/SHORT was twice as wide.  As you're (sensibly) allowed to change the formula between iterations you could change it back when difficulty rises reduce a bit.

Your goal should be that .02 range represents all difficulty-change possibilities which there's a non-trivial chance of happening and ONLY those.

I'd also suggest a table showing sample settlement values for LONG/SHORT for some difficulty changes - such as 0%, 10%, 20%, 30%, 40%, 50%.

And maybe add extra brackets into the formula to make it perfectly clear that "/100*2" does NOT mean "/(100*2)".

[woodrake]

Hi,

I'm a BTC TC moderator (username "woodtech"), but given the difficulties using the comments there to discuss a security felt it better to come to the thread.

Trust

First, Can you provide us with some assurance of your personal integrity? I do not wish to offend, but you are asking people to hand over large sums of cash and there is little evidence of who you are or your credentials. You have no trust rating on this forum and been a member for barely three months. There is no LinkedIn profile link on the securities nor any other indication of your real identity.

You said it yourself:

Third party risk: Investors have to trust BTC-TC and the issuer. Any malfunction or hack at BTC-TC or the possibility that the issuer might not be trustworthy and embezzle investors deposits could result in a total loss of the investment.

Why should we trust you? You are asking for 1000000 x 0.0204 BTC = 20,400 = $2.5m. That is a massive IPO for someone virtually unknown and hiding behind a virtual identity. I'm perhaps being a little mean but have made it my mission to raise the bar in terms of BTC TC and LTC Global securities. If I can be satisfied I will happily change my vote.

Balancing the equation

The combined dividend payout for ET.DIFF.LONG and ET.DIFF.SHORT is always 0.02.

Where is this net yield coming from? Surely the three securities should be net zero change in terms of their internal balance sheets, so there is no room for paying out net dividends?

I may have misunderstood here though; is the 0.02 transferred in from ET.DIFF.FUTURE, and if so how often does that happen?

Fees

$50k as a management fee each is a little debatable, however it is a one off which I'm less comfortable with. How do you get remunerated going forwards for running this security?

I'd be a lot happier if you were taking a smaller slice of the total invested at each difficulty change. Place a rough value on your time and estimate how long it will take you to do the administration. I'd suggest that your original 2% spread over a year (ie. 0.065% per 12 day window) indefinitely would be much better.

Also, I liked Deprived's suggestion of some tables. Would help clarify things.

Kate.

[eltopo]

Security seems fine - though only responsive in the range -50% to +50% per change.

The upper end isn't an issue but the lower end is.

I don't think anyone expects difficulty to FALL in the coming months, so allowing for it to in the payouts is inefficient.

The formula for payout on LONG is :

0.01 + ((current difficulty - last difficulty) / last difficulty) / 100 * 2

That means a payment of 0.01 if difficulty doesn't change at all - and more if it rises.  That seems wasteful in an environment where a fall just isn't likely - it means half of all capital invested is paying fees without ever being exposed to any useful risk.  In effect the shares are half the price and an additional extra amount is being tied up, shuffled around and having fees paid on it when everyone knows it'll just go back to LONG at the end.

I think if you remove the "0.01 +" and change the 2 to a 4 you'd have an instrument which sold at the same price but where the range for LONG/SHORT was twice as wide.  As you're (sensibly) allowed to change the formula between iterations you could change it back when difficulty rises reduce a bit.

Your goal should be that .02 range represents all difficulty-change possibilities which there's a non-trivial chance of happening and ONLY those.

I'd also suggest a table showing sample settlement values for LONG/SHORT for some difficulty changes - such as 0%, 10%, 20%, 30%, 40%, 50%.

And maybe add extra brackets into the formula to make it perfectly clear that "/100*2" does NOT mean "/(100*2)".

I thought about changing the factor depending on how much the difficulty changes in a period. But I wasn't sure if investors would accept a formula and contract that change every now and then. So I sticked to the 2x factor.

But it would be no problem to change it to a more "appealing" formula for the next time as difficulty isn't likely to fall.

Are there any other opinions on how this should be handled?

[woodrake]

See changes to my earlier post in case you're mid reply!

[eltopo]

Hi,

I'm a BTC TC moderator (username "woodtech"), but given the difficulties using the comments there to discuss a security felt it better to come to the thread.

Trust

First, Can you provide us with some assurance of your personal integrity? I do not wish to offend, but you are asking people to hand over large sums of cash and there is little evidence of who you are or your credentials. You have no trust rating on this forum and been a member for barely three months. There is no LinkedIn profile link on the securities nor any other indication of your real identity.

You said it yourself:

Third party risk: Investors have to trust BTC-TC and the issuer. Any malfunction or hack at BTC-TC or the possibility that the issuer might not be trustworthy and embezzle investors deposits could result in a total loss of the investment.

Why should we trust you? You are asking for 1000000 x 0.0204 BTC = 20,400 = $2.5m. That is a massive IPO for someone virtually unknown and hiding behind a virtual identity. I'm perhaps being a little mean but have made it my mission to raise the bar in terms of BTC TC and LTC Global securities. If I can be satisfied I will happily change my vote.

Balancing the equation

The combined dividend payout for ET.DIFF.LONG and ET.DIFF.SHORT is always 0.02.

Where is this net yield coming from? Surely the three securities should be net zero change in terms of their internal balance sheets, so there is no room for paying out net dividends?

I may have misunderstood here though; is the 0.02 transferred in from ET.DIFF.FUTURE, and if so how often does that happen?

Fees

$50k as a management fee each is a little debatable, however it is a one off which I'm less comfortable with. How do you get remunerated going forwards for running this security?

I'd be a lot happier if you were taking a smaller slice of the total invested at each difficulty change. Place a rough value on your time and estimate how long it will take you to do the administration. I'd suggest that your original 2% spread over a year (ie. 0.065% per 12 day window) indefinitely would be much better.

Also, I liked Deprived's suggestion of some tables. Would help clarify things.

Kate.

Hi Kate,

thanks for your feedback.

I think you misunderstood some parts of the contract. This is a future, so I'm not selling 1,000,000 "shares" (I could also set up 1,000,000,000 it would change nothing), I sell the amount investors want to buy. After a difficulty change, everything that was collected by the sales from ET.DIFF.FUTURE will be paid back to investors (minus management fee). So if I sell 100 ET.DIFF.FUTURE, I will get 100 x 0.0204 = 2.04 BTC. After the difficulty change, 2 BTC is paid out in dividends and 0.04 is the management fee. The yield for investors is coming from trading between the investors themselves (the construct is the same as the DMS assets from Deprived).

Concerning trust and my virtual identy: I asked Burnside to lock my account for withdrawal, so I can only withdraw with the help of Burnside as BTC-TC admin. This suddenly makes me as trustworthy as Burnside 

If I was able to clear up your doubt, I'd be happy if you change your vote!

[FloatesMcgoates]

I don't think woodtech knows what this security even is,

You are asking for 1000000 x 0.0204 BTC = 20,400 = $2.5m. That is a massive IPO

Is clearly an incorrect assessment.

[Rannasha]

Concerning trust and my virtual identy: I asked Burnside to lock my account for withdrawal, so I can only withdraw with the help of Burnside as BTC-TC admin. This suddenly makes me as trustworthy as Burnside 

I'm not sure how flexible Burnsides account-locking-features are, but for this asset to function properly, you have to be able to transfer units (the FUTURE -> LONG + SHORT transfer), so the account should not be locked out of this. But if the account can transfer units out, there is still the possibility for malicious transfers (i.e. transfer 10000 units to a secondary account, sell everything on the open market and withdraw the funds), so a simple withdrawal-lock isn't enough to completely remove the counterparty risk.

I make no claims about whether eltopo is trustworthy or not, simply that the proposed mechanism of an account lock is insufficient to make the question of his trustworthiness moot.

Then again, many securities/funds/whatevers on BTC-TC offer at best relatively minimal guarantees to the trustworthiness of the issuer, so this security is not any different in that regard.

[eltopo]

Burnside has not answered yet. I suppose the lock has only an affect on BTC withdrawals out of BTC-TC, and not transfers or other things needed for operation.

[Progressive]

The lock needs to be on transfer of BTC also - otherwise you could transfer the BTC to another account and withdraw from it.
If burnside locks both withdrawals and BTC transfers then I will have no trust issues (Rannasha has a good point).

The points mentioned by Deprived are very interesting, I didn't think about this:

That means a payment of 0.01 if difficulty doesn't change at all - and more if it rises.  That seems wasteful in an environment where a fall just isn't likely - it means half of all capital invested is paying fees without ever being exposed to any useful risk.  In effect the shares are half the price and an additional extra amount is being tied up, shuffled around and having fees paid on it when everyone knows it'll just go back to LONG at the end.

I think if you remove the "0.01 +" and change the 2 to a 4 you'd have an instrument which sold at the same price but where the range for LONG/SHORT was twice as wide.  As you're (sensibly) allowed to change the formula between iterations you could change it back when difficulty rises reduce a bit.

Eltopo, what about the change Deprived proposed?

EDIT: Eltopo, would you mind disclosing your identity? Either publicly or to some trusted party (burnside, John K.).

[eltopo]

The lock needs to be on transfer of BTC also - otherwise you could transfer the BTC to another account and withdraw from it.
If burnside locks both withdrawals and BTC transfers then I will have no trust issues (Rannasha has a good point).

I asked burnside to answer in this thread, let's wait what he tells us.

The points mentioned by Deprived are very interesting, I didn't think about this:

That means a payment of 0.01 if difficulty doesn't change at all - and more if it rises.  That seems wasteful in an environment where a fall just isn't likely - it means half of all capital invested is paying fees without ever being exposed to any useful risk.  In effect the shares are half the price and an additional extra amount is being tied up, shuffled around and having fees paid on it when everyone knows it'll just go back to LONG at the end.

I think if you remove the "0.01 +" and change the 2 to a 4 you'd have an instrument which sold at the same price but where the range for LONG/SHORT was twice as wide.  As you're (sensibly) allowed to change the formula between iterations you could change it back when difficulty rises reduce a bit.

Eltopo, what about the change Deprived proposed?

I answered Deprived's post above and asked for more opinions about this. I don't have a problem with this change. If investors like it, I like it too 

EDIT: Eltopo, would you mind disclosing your identity? Either publicly or to some trusted party (burnside, John K.).

I would disclose my identity to Burnside if that helps.

[woodrake]

I think you misunderstood some parts of the contract. This is a future, so I'm not selling 1,000,000 "shares" (I could also set up 1,000,000,000 it would change nothing), I sell the amount investors want to buy.

Hmm, yes I see what you mean. It is indeed not fair to compare you to a stock/company IPO. Also, the investors / market can make up their own minds how much to trust you by dipping their toes in.

I do perhaps sometimes overly fret about people doing a running with the funds; it would be so easy for this to happen, especially on this sort of security where there is little to show for the investment such as work to date, a functioning Web site, purchase receipts or a portfolio manifest.

I am conflicted in my position as a moderator since I don't know how much I should be doing to try protect the investors.

After a difficulty change, everything that was collected by the sales from ET.DIFF.FUTURE will be paid back to investors (minus management fee). So if I sell 100 ET.DIFF.FUTURE, I will get 100 x 0.0204 = 2.04 BTC. After the difficulty change, 2 BTC is paid out in dividends and 0.04 is the management fee. The yield for investors is coming from trading between the investors themselves (the construct is the same as the DMS assets from Deprived).

Thank you for the patient explanation, I understand now. I think I rushed reading the original contract! To summarise so that I can be sure I have understood, say I want to bet that difficulty is going to be higher than expected (this is a realistic scenario in my case since I'm heavily in mining and may wish to hedge my risk):

• I purchase 500 ET.DIFF.FUTURE @ 0.02 BTC (10 BTC total)
• I transfer all 500 ET.DIFF.FUTURE back to you
• You issue me with 250 ET.DIFF.SHORT and 250 ET.DIFF.LONG
• I sell all my ET.DIFF.SHORT and perhaps purchase some additional ET.DIFF.LONG
• At the end of the 12 day window difficulty has risen by 25%; ET.DIFF.LONG pays out a dividend of 0.01025 then is bought back at 0.0

First, step 5 appears wrong. I calculated that using your formula, but it is actuall ambiguous due to precedence:

Code:

Dividend = 0.01 + ((current difficulty - last difficulty) / last difficulty) / 100 * 2

I think you meant:

Code:

Dividend = 0.01 + ( ( ( ( current difficulty - last difficulty ) / last difficulty ) / 100 ) * 2 )

Second, I agree with Deprived that it does not make sense to assume difficulty might fall. I think you should re-engineer this so that the SHORT/LONG equilibrium mid-point is set at what the market, on average, thinks is going to happen. In the event of zero difficulty change the LONG divi should be 0.0 and the SHORT 0.02. I'm not sure how best to achieve that though!

Concerning trust and my virtual identy: I asked Burnside to lock my account for withdrawal, so I can only withdraw with the help of Burnside as BTC-TC admin. This suddenly makes me as trustworthy as Burnside 

If I was able to clear up your doubt, I'd be happy if you change your vote!

That's a commendable move, well done! As Progressive said though the lock would have to be on internal transfers also. This would not prevent you from embezzling funds of course (you could sell a personal account shares in something at a non-market price or move options around), but it would be a positive step in terms of engendering trust.

I think that in order for Burnside to agree you would have to make a commitment to him that you'd never make a withdrawal without a motion or something like that, to reduce his workload.

Regardless, I'm much happier about this overall now. I've moved to abstain while you get back to me.

Kate.

[Rannasha]

• I purchase 500 ET.DIFF.FUTURE @ 0.02 BTC (10 BTC total)
• I transfer all 500 ET.DIFF.FUTURE back to you
• You issue me with 250 ET.DIFF.SHORT and 250 ET.DIFF.LONG
• I sell all my ET.DIFF.SHORT and perhaps purchase some additional ET.DIFF.LONG
• At the end of the 12 day window difficulty has risen by 25%; ET.DIFF.LONG pays out a dividend of 0.01025 then is bought back at 0.0

Almost, but there are still some inaccuracies.

• I purchase 500 ET.DIFF.FUTURE @ 0.0204 BTC (10.2 BTC total) (2% mgmt fee)
• I transfer all 500 ET.DIFF.FUTURE back to you
• You issue me with 500 ET.DIFF.SHORT and 500 ET.DIFF.LONG (each FUTURE corresponds to 1 unit of both SHORT and LONG)
• I sell all my ET.DIFF.SHORT and perhaps purchase some additional ET.DIFF.LONG
• At the end of the 12 day window difficulty has risen by 25%; ET.DIFF.LONG pays out a dividend of 0.015 then is bought back at 0.0 (Payout with current formula is 0.01 + twice the fractional increase divided by 100)

First, step 5 appears wrong. I calculated that using your formula, but it is actuall ambiguous due to precedence:

Code:

Dividend = 0.01 + ((current difficulty - last difficulty) / last difficulty) / 100 * 2

I think you meant:

Code:

Dividend = 0.01 + ( ( ( ( current difficulty - last difficulty ) / last difficulty ) / 100 ) * 2 )

This might be just me being a pedantic mathematician, but the two expressions are equivalent. Division and multiplication have equal priority, so these operations are executed in order from left to right. So x / 100 * 2 is computed by first dividing and then multiplying. No brackets needed.

Second, I agree with Deprived that it does not make sense to assume difficulty might fall. I think you should re-engineer this so that the SHORT/LONG equilibrium mid-point is set at what the market, on average, thinks is going to happen. In the event of zero difficulty change the LONG divi should be 0.0 and the SHORT 0.02. I'm not sure how best to achieve that though!

Dividend = (C - L) / L / 100 * 4
This caps out at a 50% increase (all money goes to holders of LONG) and at difficulty stagnation (all money goes to holders of SHORT).

[eltopo]

Hi Kate,

please read what Rannasha wrote, he got it 100% right (Thanks for answering, Rannasha).

It seems most investors prefer a formula with a cap at 0% at the bottom and 50% at the top (like Deprived suggests). The formula would look like this:

Dividend ET.DIFF.LONG = ((current difficulty - last difficulty) / last difficulty) / 100 * 4

The dividend formula for ET.DIFF.SHORT would stay the same (0.02 - dividend for ET.DIFF.LONG).

I will change the contract this way. As soon as difficulty stops rising, I will change it back to a formula that takes descreasing difficulty into account (this contract change will of course be announced on BTC-TC and bitcointalk forums at least 3 days before they take effect).

Concerning account lock, we'll have to wait on burnside and check if it's possible to lock withdrawal and transfer-out of BTC, but not locking transfer of assets as it is needed for operation.

[eltopo]

Contract has been updated in the first post and on BTC-TC (formula changed on dividend payments).

The third post has been updated with a payout scheme of dividends.

[Deprived]

I'm not sure an account lock can be done in such a way as to stop fraud anyway.

People who think blocking withdrawals/transfers of cash prevents it are wrong.  Eltopo could just buy shares then transfer those to a different account.  Funds could even be transferred using his own shares.

I don't think it's worth the bother as all it does is add a few minutes to the time needed to steal (and some cost making it less efficient) - it in no way prevents it.

[ThickAsThieves]

I'm not sure an account lock can be done in such a way as to stop fraud anyway.

People who think blocking withdrawals/transfers of cash prevents it are wrong.  Eltopo could just buy shares then transfer those to a different account.  Funds could even be transferred using his own shares.

I don't think it's worth the bother as all it does is add a few minutes to the time needed to steal (and some cost making it less efficient) - it in no way prevents it.

True, but if Burnside is willing to implement the concept of having accounts with varying degrees of lockdown, it would make securities like this be forced to choose the levels of lockdown appropriate to their cause, and thus prevent certain problems, and remove certain trust requirements.

Of course I'm not sure what amount of lockdown is achievable without fully scripting all aspects, and at that point Burnside might as well just offer sets like these himself and take issuers out of the picture.

[Deprived]

I'm not sure an account lock can be done in such a way as to stop fraud anyway.

People who think blocking withdrawals/transfers of cash prevents it are wrong.  Eltopo could just buy shares then transfer those to a different account.  Funds could even be transferred using his own shares.

I don't think it's worth the bother as all it does is add a few minutes to the time needed to steal (and some cost making it less efficient) - it in no way prevents it.

True, but if Burnside is willing to implement the concept of having accounts with varying degrees of lockdown, it would make securities like this be forced to choose the levels of lockdown appropriate to their cause, and thus prevent certain problems, and remove certain trust requirements.

Of course I'm not sure what amount of lockdown is achievable without fully scripting all aspects, and at that point Burnside might as well just offer sets like these himself and take issuers out of the picture.

Reason I don't like it is that it gives a false sense of security.  Scammers can still scam just they get something to hide behind that the naive might believe prevents them from scamming.

With securities like this one it's pretty much impossible to prevent scamming at an exchange level - though you CAN make it more obvious when they're trying.  One easy way to steal all the funds would be:

1.  Do forced buyback on all shares at 1 satoshi.
2.  Transfer some shares out to alt account
3.  Do dividend payment which gives out all funds to the alt.

Transfer of shares, dividends and forced buybacks are necessary for operation of the securities - so can't be blocked.

I can immediately think of less obvious ways - but fact is that without logic in place to check whether transactions are valid (or manual approval of them) scamming/theft can't be prevented.  And if it can't be prevented then there's no benefit in pretending that it has - as that just makes life easier for scammers by giving them a means to obtain credibility without actually stopping or majorly hindering them.

I'm all for anyhting which stops scamming/theft - but it needs to actually do so not just make it slightly harder.  Or the downside (making it easier for scammer to gain trust) outweighs the benefits (making it slightly easier for honest people with little reputation to gain investment).

[Deprived]

I see this security as a pretty low-risk one (from a scamming perspective) anyway.

It's clear where issuer's profit comes from - 2% management on turnover.

As there's a full reset every difficulty change, the most he could steal would be the cash from one cycle.

If investment increases each cycle (as trust in him grows) then he'd make more from being honest (but over a longer time-frame) than from stealing.  Which is the single biggest factor I look at when assessing CP risk - if someone's making a nice profit by behaving honestly then there's (in general - there ARE exceptions) no rational reason for them to steal even if they were dishonest by nature.

I don't know eltopo particularly well - other than that he's been an active participant in DMS (both in the thread and in trading - I see his name plenty of times on transfers).  But at least I'm confident he understands how to run the security, how mining difficulty works and isn't penniless.  Not all of which are true for some of the mining operations being IPOed recently.

[eltopo]

Thanks for the words, Deprived.

While I understand the caution investors have against a new security and issuer (I do the same with other securities), in the end it comes all down to upfront trust the investors need to have when investing in new securities. If I don't trust a new issuer, I stand at the sideline and wait a few weeks until the issuer proved to be trustworthy and capable of managing the security. Especially as there are no real options to prevent a scammer from running with the funds.

I just could say that I'll try my very best to manage the asset (and I will), but so would a scammer say, too.