Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - binggo

Pages: 1 ... 99 100 101 102 103 104 105 [106] 107 108 109 110 111 112 113 ... 159
1576
      要么给喂价设定一个底线,既不得低于(内盘价*鼓鼓充值费率)

这个可以有。

早就是这么规划的。去年就是。后来实施的时候不知道为啥你改成了内外盘价格比做折价了。
原来所说的就是鼓鼓充值费率进行修正。

不过,我还是想再进一步,干脆也别充值费率了,干脆就喂价必须大于内盘买单价得了。

不建议, 大户会内盘抵押出大量锚定资产购买BTS来把内盘价格拉高,虽然会有外盘来砸盘,但是内盘喂价也会升的很快.

虽然这样的大户不多,但是几个大户一联合,外加贴线抵押的助推也不可小看其带来的危害.

修正:最好是取(中值喂价+内盘价格)的均价,最终喂价不得大于中值喂价的102%, 不得小于中值喂价.

见证人还是按照现在的方式喂价,系统也按现在的方式取中值价格,然后计算出这个最终的价格作为喂价。

最终喂价可以偏离中值喂价,但是不能偏离太多,高出中值喂价的部分做为见证人之间各种汇率换算的差额补偿,但是不能低于中值喂价。

这样猴子哥就可以尽情抵押拉盘.

1577
After BSIP42 was unvoted, magicwallet just use a "medium" price for months (multi sources include BitShares blockchain). Same as clockwork, there is no USDT in any sources. In these days, as BTS price in BitShares are higher than **many** external exchanges, it seems the feed price is low. Now magicwallet changed to use a "mean" price, it is slightly higher. However, there is no perfect strategy. As many external exchanges has faked volume, the following equation may be  better:

   mean( medium price of external sources , bts price in BitShares blockchain).

For these days, the feed price seems low, but it is health in my opinion: It is a warning sign to not just use MPA( which is derived from BTS) to bring BTS up. On the other hand, it is hard to be manipulated to eat the margin call to make profit as the market price is higher.

So the wittness can use the feed price to control the market?en?so interesting.

Now i think @ljk424 said is right: Feed price is the cancer of  BTS, the feed price which was provided by people is the cancer of BTS, anthropogenic factor and politics in there.

Just like the @ljk424 said.

1578
Please remove the vote from gdex-wittness until he correct his feed script.

1579
这种喂价就是系统的最大毒瘤,没有之一!人为的制造连环爆仓,自我阉割式的砸盘,这么多年再烂的币都涨上去了,为啥bts还这熊样?内盘已经比很多外盘都大了,结果内盘自己却一点没有定价权。为什么要跟着外盘走呢?刚才跌的时候,内盘根本就没怎么跌啊,硬是给喂价砸下去的,完全是自杀式的做法。假如有一天内盘成交量最大了,难道还是要看外盘脸色?我们见证人自己都不支持自己的内盘,bts怎么可能发展呢?能不能给内盘一点机会抵抗空头?
      要么给喂价设定一个底线,既不得低于(内盘价*鼓鼓充值费率);要么采用最高价原则;要么搞个指数平衡一下。方法这么多,能不能行动一下啊!

(内盘价*鼓鼓充值费率)不能作为喂价参考,其影响因素太多,太杂,人为干涉因素太深。

整个系统中喂价是人为因素最大的一环,靠见证人来搞好喂价现在基本可以死了心了。

参数要加在中值喂价上,不是见证人想出的各种奇妙点子加在自己的喂价上!!!

的确如某人所描述的:懒,散,慢,理事会不同步更新自己的投票方向,见证人不更新自己的工作进展与喂价情况,无人监督,太多的政治。

太多人喜欢因循守旧。

都在比谁比谁喂的价低,这才是蔚为壮观。

另外: 实在不想评论 gdex-wittness 上窜下跳猴子一样的喂价方式,连基本的平滑性都没有。

1580




Pleas don't use the bitCNY premium/discount in the feed price, as have too many factors to influence the bitCNY premium/discount.

witness.hiblockchain   0.4744
gdex-witness


explain?

1581
价格涨,喂价装死,价格跌,喂价领跌,致敬某些这帮XXXX的见证人!!!!!

设计机制的漏洞,太多人为的因素,不能怪见证人。

如果把这个漏洞堵上,也算是去了一大顽疾!

1582
Stakeholder Proposals / Re: [Poll] BSIP59:Reduce MCR of bitCNY to 1.6
« on: April 08, 2019, 09:42:51 am »
When do u think the right time to reduce MCR?

In arround 2 months.There are many indicators we will see a last final bigger downtrend (third bottom) before getting the real uptrend.
Reduce of MCR should only happen in real uptrend to avoid getting negative effects in a downtrend.
Agree,But the market is hard to predict and the right time is hard to grasp.

The settlement, we can use it to let the market to control themself.

The settlement offset, the settlement delay, the TARGET COLLATERAL RATIO to control the MCR,even set the MCR=1.5 or 1.35.


1583
Guys, please calm down ...

Due to the use of "MEDIAN" in deriving the feed price on chain, you will **ALWAYS** .. let me repeat that: **ALWAYS**
find half the witnesses produce a price that is *above* the price feed and the other half to be *below* the price feed!!

I see, but the" MEDIAN" price is not a fair price, i don't think the price which come from the BTS/BTC price is a true fiat pirce, too many exchang and charge in there,  BTS-BTC-USDT(now6.93)-USD(6.7197)-CNY.

and i want to know what happened to the gdex-wittness, he is the highest in the last days, now he is the lowest!!!

If we can't make a professional feed price,how we can make a StableCoin?

1584
我们连一个正常的喂价都锚不住?

现在这种喂价对比真的是让人无言以对!!!

witness.yao   0.5009
zapata42-witness   0.5009
in.abit   0.4997
xeldal   0.4981
bhuz   0.498
openledger-dc   0.4978
roelandp   0.4977
delegate-zhaomu   0.4964
sahkan-bitshares   0.4936
btspp-witness   0.4931

bangzi   0.4922
witness.hiblockchain   0.4904
fox   0.4897
xn-delegate   0.4895
verbaltech2   0.4888
clockwork   0.4887
delegate-1.lafona   0.4885

magicwallet.witness   0.4877
blckchnd   0.4876
witness.still   0.4875
gdex-witness   0.4863

第一种方法得出的修正喂价:
0.00639503
0.495347634

第二种参数得出的修正喂价:
0.007861318
0.496069184

喂价低于内盘价格,取两者均值:0.4961


内盘自己没崩,被外盘给出的喂价带崩,是不是很悲哀?

当内盘价格大于或等于喂价时,喂价权重里加上至少20%(不,应该是50%)的内盘深度加权价格;
当内盘价格小于或等于喂价时,喂价权重里加上至少20%(不,应该是50%)的内盘深度加权价格×MSSR; 现在MSSR=102%的情况下,完全可以加上。

1585


so,what's wrong with the feed price? what's wrong with the wittness?

Squeezing the market by the feed price?

DEX PRICE: 0.5

witness.yao   0.5009
zapata42-witness   0.5009
in.abit   0.4997
xeldal   0.4981
bhuz   0.498
openledger-dc   0.4978
roelandp   0.4977
delegate-zhaomu   0.4964
sahkan-bitshares   0.4936
btspp-witness   0.4931
bangzi   0.4922
witness.hiblockchain   0.4904
fox   0.4897
xn-delegate   0.4895
verbaltech2   0.4888
clockwork   0.4887
delegate-1.lafona   0.4885
magicwallet.witness   0.4877
blckchnd   0.4876
witness.still   0.4875
gdex-witness   0.4863


1586
Make the settlement offset vary with CR;


K = Modification Factors(e.g 0.02)
x = Modification Factors(e.g. 3)

Force settlement offset= K*(CR/MCR)^x

    CR above the MCR will get more settlement offset;
    e.g.
    when MCR=160 K=0.02,x=2,the 200% collateral ratio will get 3.1% Force settlement offset.
    when MCR=150 K=0.02,x=3,the 200% collateral ratio will get 3.9% Force settlement offset.

Code: [Select]
# -- coding:utf-8 --

import matplotlib.pyplot as plt
import numpy as np

K=0.01
X=4
CR=np.arange(100,300)
MCR=160
offset=K*(CR/MCR)**X

plt.xlabel("CR",fontsize=15)
plt.ylabel("offset",fontsize=15)

plt.plot(CR,offset)
plt.show()


    If we need to set a Maximum settlement offset? 5% or other parameters;
    The settlement Delay above MCR still be 24 hours.

https://github.com/bitshares/bsips/issues/160

1587
Stakeholder Proposals / Re: [Poll] BSIP59:Reduce MSSR of bitUSD
« on: April 07, 2019, 08:06:19 am »
Support :1.14.179   Poll - BSIP59 - Reduce MSSR of bitUSD to 1.02


1588
有MSSR的存在,内盘价格在价格下跌的时候很难反应实际法币价格,往往会大幅低于实际法币价格,直接采内盘价格做喂价不合适。

而影响入金费率的因素又太多,入金费率也不适合添加到喂价参数里。

喂价偏低还是主要是国内与国外见证人之间的汇率转换不一样,在国外社区见证人比较多的情况下,中值喂价就会偏低与实际法币价格。

见证人之间的汇率是很难一致,谁也不好说谁就是错的,毕竟CNY的汇率还有离岸与在岸之别,USDT与USD之间也有汇率差,BTC再转到USD。

1. 怎么设置一个汇率差别参数来平衡见证人之间的汇率差,尽量降低本币与外币之间汇兑转换之间导致的偏差?

2. 怎么才能使内盘的价格也能做喂价采集?而且也要预防内盘崩盘时内盘作为喂价采集导致连环平仓?

喂价比较法?外盘价格与内盘价格哪一个高就用哪一个?

或者外盘价格大于内盘价格时,外盘价格做喂价?外盘价格低于内盘价格时,取两者的平均值做喂价? 我感觉这样折中一下最好.

或者我们设一个参数 Z=绝对值(喂价-内盘时间深度加权价格)/(喂价+内盘时间深度加权价格),  最终喂价=喂价×(1+Z)?
这个参数在没有大量爆仓单累积的时候还好说, 有大量爆仓单的时候, 最终喂价会高于喂价与内盘价格, 很多该爆的爆不了, 等外盘回暖的时候, 内盘拉盘会出现喂价降低, 又增加爆仓单的弊端.
这种参数有个弊端就是价差过大就会导致最终数值偏差过大, 这种参数怎么设置都有问题,
当内盘价格大于外盘价格的时候, 内盘价格不低于外盘5%价格的时候,这种参数还好用.

就跟喂价分歧指数一样, 有三个见证人喂高价,就会导致最终喂价偏离过大.


说一些题外话:锚定资产是如何实现锚定的? 顺序是怎么排的?

现在是这种顺序:

bts的外部市场价值(也是实际法币价格)——内盘喂价——1.75倍的BTS抵押产出锚定资产——锚定法币。

又因为bts自带交易系统,可以实现bts与锚定资产的内部流通交易,实现变相的自我锚定,那么假如在没有外部市场存在的情况下,是否能够实现自我锚定?

如果要实现这种自我锚定的前提条件是什么?我认为需要以下:
1. bts要体现自身价值需要与外部有联系出口,一种是承兑,一种是大品种代币的跨链比如与BTC;
2. 抵押借贷不能以打折形式投向市场;
3. 对锚定资产生产者要形成适当的保护而不过量;
4. 累积的锚定风险要得到合理的释放;
5. 锚定资产持有者也要有一定的约束,虽然有贬值风险,但是以历次锚定资产全局清算的结果来看,锚定资产持有者并不担心全局清算;

这在BSIP18规则的约束下,抵押者所承担的风险过大,当然你也可以说他们是咎由自取,不能自己控制风险,但是别忘了一点,锚定资产的流动性是由他们提供的,没有他们,锚定资产就是空话,当系统整体债仓的负债率达到40%时,任何锚定资产的持有者都是借款人,都需要去偿还将要失去价值的锚定资产,来置换回价值抵押物,抵押物才是真正的价值黄金,这对锚定资产持有人来说并没有太大损失。

1.场外交易体现bts实际价值;
2.承兑体现bitcny实际价值,间接体现bts实际价值;
3.内部交易所的其它交易对来体现bts的实际价值;

bts的实际价值是实现抵押借贷的基础;

论喂价存在的意义,
个人的看法是喂价存在有两个意义:

1.当此交易对流动性差时,提供指导性价格;

2.在平仓时起相对保护性的作用,尤其是在流动性差及高MSSR的情况下,在流动性好高MSSR的情况下也起关键作用。

建立在BTS上的交易所规模不大或者跨链不成熟规模也不大的话,现阶段喂价也没有很好的调整办法。

那我们就需要尽可能的去削弱喂价对整体抵押的影响,我的意见依然是依靠强清这个机制来削弱喂价,MSSR导致的平仓效应。

这样我们可以降低MCR,用强清参数(强清延迟,强清补偿,强制强清目标抵押率)来实现动态控制MCR,在平时就释放风险。

强制平仓在市场下行是才是对整个内盘压制最大的,让市场情绪缓和个半个小时,一个小时,结果完全不同。

像期货市场是系统来接管债仓在市场中寻找最优价格来平仓位。而利用强清我们就可以不用设计系统接管债仓平仓的这个功能,而是让市场自己寻找利润机会利用强清在市场中寻找最优价格平仓,这难道不是更好的化解风险的办法?

1589
这似乎是旧版本了吧

1590
General Discussion / Re: Balance the exchange of feed price
« on: April 05, 2019, 07:54:00 am »
2 witness collude to feed 50 : Z becomes ~0.9 -> feed almost doubles

en, this is a problem, maybe we change the

Z=(The sum of the five effective feed price highter than median of witness feeds - The sum of the five effective feed  price lower than median of witness feeds)/(The sum of the five effective feed price highter than median of witness feeds+The sum of the five effective feed price lower than median of witness feeds);

and give it a limits:max 2%

maybe have some good ways.

Pages: 1 ... 99 100 101 102 103 104 105 [106] 107 108 109 110 111 112 113 ... 159