c*********t
发帖数: 2341 | 1 原来公式里面的i和g就是预期了，后面的delta(P/E)其实还是反应了i和g的变化，其实
就是delta(i)加delta(g)
除非原来的i和g是常数不随时间变化，否则岂不是把delta(i)和delta(g)算了两遍？ |
|
|
|
b*****e
发帖数: 1125 | 4 你在考CFAL2?
i and g are expected inflation rate and real earning growth rate. They are
independent of p/e ratio since p/e could not change with earnings
continuously growing. |
|
c*********t
发帖数: 2341 | 5 ft，我考CFA level2是n年以前的事了
我的意思是说p/e ratio之所以会发生变化就是因为市场对inflation和growth的预期发
生了变化
所以delta(p/e)实际上就是delta(i)+delta(g) |
|
b*****e
发帖数: 1125 | 6 Expected Returns= Div1 /P0 + i + g - ΔS + Δ (P/E)
Div1 = dividend in next period (period 1 assuming current t=0)
P0 = current price (price at time 0)
i= expected inflation rate
g= real growth rate in earnings (note that by adding real growth and
inflation, this is basically identical to just adding nominal growth)
ΔS= changes in shares outstanding (i.e. increases in shares outstanding
decrease expected returns)
Δ(P/E)= changes in P/E ratio (positive relationship between changes in P/e
and expect |
|
c*********t
发帖数: 2341 | 7 假设这个公式用在t=1的时候，那么
Δ (P/E)应该是t=1的时候的P/E减去t=0的时候的P/E对吧
其反应的就是市场在从t0到t1这段时间对inflation和growth预期的变化
e |
|
b*****e
发帖数: 1125 | 8 That's right. But it means Δ(P/E) depends on Δi and Δg, it has nothing to
do with E(i) and E(g) at time t0.
I think this formula used at t0, E(i) and E(g) are both determined at t0 and
Δ(P/E) is an expected change of P/E during t0 and t1. |
|
c*********t
发帖数: 2341 | 9 问题是这个E(i) and E(g)是t1的时候的啊
to
and |
|
b*****e
发帖数: 1125 | 10 OMG. Have you ever learned probability?
Look at the left hand side of the formula, its expected return which is at
t0. So anything on the right hand side without post-fix should be value at
t0.
BTW expected value are typically long term expectation which should be in
decades. So whether you are at t0 or t1, E(i) and E(g) should be the same. |
|
c*********t
发帖数: 2341 | 11 ft，前面我都说了按照t=1算
如果公式左边是t=0那右边也是t=0
还是一样的把从t=-1到t=0这段时间delta(i)和delta(g)算了两遍 |
|
|
b*****e
发帖数: 1125 | 13 Please read the following formula again and tell me what time is on the lhs
and what time is i and g locates at?
Expected Returns= Div1 /P0 + i + g - ΔS + Δ (P/E)
Div1 = dividend in next period (period 1 assuming current t=0)
P0 = current price (price at time 0) |
|
|
b*****e
发帖数: 1125 | 15 Please read the following formula again and tell me what time is on the lhs
and what time is i and g locates at?
Expected Returns= Div1 /P0 + i + g - ΔS + Δ (P/E)
Div1 = dividend in next period (period 1 assuming current t=0)
P0 = current price (price at time 0) |
|
c*********t
发帖数: 2341 | 16 你还是没明白我的意思，i和g是在t=0这个点上对未来长期的通胀和增长的预期
而 Δ (P/E)反应的是市场P/E值从t=-1到t=0这段时间的变化
这么说没有问题吧
lhs |
|
b*****e
发帖数: 1125 | 17 My understanding is: Δ(P/E) = p0/E1 - p1/E0
but E(g+i) <> E(E1-E0), its a long term earning growth forecasting, so it
needs adjustment. |
|
c*********t
发帖数: 2341 | 18 Δ(P/E) = p0/E1 - p1/E0
这个不对吧
另外我也没说E(g+i) = E(E1-E0)，我是说E(g+i) 包含E(E1-E0)
所以多加一块是画蛇添足了
总而言之我觉得这个公式本身有缺陷 |
|
b*****e
发帖数: 1125 | 19 typo: Δ(P/E) = p0/E1 - p(-1)/E0
E(g+i) = E(E1-E0) by defition
This is an empirical formula and without suprise has shortage. |
|
|
v*******n
发帖数: 8995 | 21 也就是 grinold fundamental law
这个公式很好地描述了我的表现
简单来说就是预测挺准，下得还多 |
|
c*******e
发帖数: 150 | 22 two recs
R. Grinold
Campbell, Lo & MacKinlay
I'm not saying you must finish these two books. But some decent familiarity
with these is analogous to knowing J. Hull and Shreve vol II for a
professional in the derivatives space. (I wouldn't say either reaches the
equivalent to Karatzas & Shreve) |
|
r**a
发帖数: 536 | 23 那位仁兄给推荐几本经典buy side quant必读书呀？多谢了。
我现在知道的有两本
1. The Econometrics of Financial Markets by Campbell
2. Active Portfolio Management by Grinold |
|
r**a
发帖数: 536 | 24 First, thanks for the useful comments. Actually,I am just a newbie for the
buy-side. Currently I only know the sell side stuff, and want to learn
something in the buy side part, especially want to understand the
statistical arbitrage and algorithm trading.
U know, for sell side quants, there are lots of good books to read and lots
of threads to tell u how to become a sell-side quant step by step starting
from the very beginning, e.g. the list suggested by Mark Joshi . But for me,
I never saw a b... 阅读全帖 |
|
c*******e
发帖数: 150 | 25 PDT 不喜欢 "拿到citadel的大牛" 是人家口味比较特殊，巨喜欢直接从学校招 职业小
处男小处女，改嫁的一律不要，看到实习经历的也倒胃口。学校的朋友们拿到和据掉
PDT的例子也见了一大把了，还很少听说谁被PDT lateral hired 的。人家喜欢的就是童
男童女。Peter Muller 武功源自 Barra，特别是Richard Grinold。 这种 style 的
investment 在最近3年的 US equity mkt 是个非常困难的时期。当然 Muller 门派下
面的很少和外帮的人员流动，能够in-house 发掘并且高度保密一些novel的 alpha
factors 而把对手甩开也是完全有可能的。不了解他们最近的 perf 不好发言
publication. |
|
c*******e
发帖数: 150 | 26 Larry Harris 可以用来扫盲/入门教育
如果还在学校里有大量的休闲的时间，Grinold & Khan 的 bible 当然也是打内功基础
的好书
之后看你 trade 什么方面的 asset class/venue 和 trading styles (e.g. stat-arb
, event-driven, trend-following, factor-based alphas, Options AMM, etc) 每一
门武功 (style) 和兵器 (instrument) 都非常地不同，师傅带进门以后修行就考自己
了。实战里面再不断提高，倒也没有什么统一的书本可以参考的 |
|
c*******e
发帖数: 150 | 27 是的，buy-side 量化交易中 常说的 Grinold & Kahn 就是指这本书。他对于
量化基金行业的普适地位不亚于 John Hull 对于sell-side 衍生品行业中的
普适地位
基础
每一
自己 |
|
