初中取消物理课了吗:金融市场的动量效应：为什么牛顿错了

为什么牛顿错了

Theory says that the past performance of share prices is no guide to the future. Practice says otherwise

有理论讲，人们无法根据一支股票的历史价格预测其未来的走势。实践中却恰恰相反。

WHAT goes up must come down. It is natural to assume that the law of gravity should also apply in financial markets. After all, isn’t the oldest piece of investment advice to buy low and sell high? But in 2010 European investors would have prospered by following a different rule. Anyone who bought the best-performing stocks of the previous year would have enjoyed returns more than 12 percentage points higher than someone who bought 2009’s worst performers.

有升就有降，有高就有低。人们很自然地认为重力法则也适用于股票市场。不管怎么说，最传统的投资建议不就是低买高卖吗？但2010年欧洲赚钱的投资者遵循的却是却另一套法则。在这一年买进2009年表现最优的股票，比买进2009年表现最差的股票要享受高出12%的收益。

This was not unusual. Since the 1980s academic studies have repeatedly shown that, on average, shares that have performed well in the recent past continue to do so for some time. Longer-term studies have confirmed that this “momentum” effect has been observable for much of the past century. Nor is the phenomenon confined to the stockmarket. Commodity prices and currencies are remarkably persistent, rising or falling for long periods.

这种情况并不独特。自从20世纪80年代起，学术研究上就不断发现，平均来讲，在过去较近一段时间表现好的股票会在未来某段时间内保持好势头。而更长时间跨度的研究则证实，这样一种动量效应在过去一百年中多数时候都可观察到。而且这种现象并不只局限在股票市场。商品价格和货币价格的走势也具有持续性，无论升势或降势都是长期的。

The momentum effect drives a juggernaut through one of the tenets of finance theory, the efficient-market hypothesis. In its strongest form this states that past price movements should give no useful information about the future. Investors should have no logical reason to have preferred the winners of 2009 to the losers; both should be fairly priced already.

动量效应对金融理论中的有效市场假说是一个很大的冲击。有效市场假说中最典型的理论认为，人们无法根据股票的历史价格预测其未来的走势。因此，投资者没道理看好2009年表现好的股票而轻视这一年表现差的股票；因为所有股票都已经是合理估值的。

Markets do throw up occasional anomalies异常,反常,不规则—for instance, the outperformance of shares in January or their poor performance in the summer months—that may be too small or unreliable to exploit. But the momentum effect is huge. Elroy Dimson, Paul Marsh and Mike Staunton of the London Business School (LBS) looked at the largest 100 stocks in the British market since 1900. They calculated the return from buying the 20 best performers over the past 12 months and then holding them, rebalancing the portfolio every month.

但市场偶尔会有不规则现象出现——例如股票在一月份表现极佳，或在夏季的几个月里表现很差。但这样的现象对投资来说不是几率太小就是不稳定，因此难以利用。而动量效应却是巨大的。伦敦商学院三位学者Elroy Dimson, Paul Marsh和Mike Staunton研究了1900年英国股市中100支最大的股票，如果一个人买入先前12个月中表现最好的20支股票，对其持有并每月做些配置上的调整，经他们计算，由此而得的年收益率比买进之前12个月中表现最差的股票要高出10.3%。

This produced an annual average of 10.3 percentage points more than a strategy of buying the previous 12 months’ worst performers. An investment of ￡1 in 1900 would have grown into ￡2.3m by the end of 2009; the same sum invested in the losers would have turned into just ￡49 (see chart 1).

使用这种策略，1900年1英镑的投资到2009年已增长为230万英镑，而买进表现低迷的股票经过这109年所得回报仅为49英镑。

Messrs Dimson, Marsh and Staunton applied a similar approach to 19 markets across the world and found a significant momentum effect in 18 of them, dating back to 1926 in America and 1975 in larger European markets. A study by AQR Capital Management, a hedge fund, found that the American stocks with the best momentum outperformed those with the worst by more than ten percentage points a year between 1927 and 2010 (see chart 2). AQR has set up a series of funds that attempt to exploit the momentum anomaly.

Messrs Dimson, Marsh 和Staunton三位学者用相同的方法研究了自1926年至1975年从美国到欧洲的19个主要股票市场，在18个股市中都发现了显著的动量效应。而由一家对冲基金AQR 资金管理公司所做的调查发现， 1927至2010年间，美国动量最好的股票比动量最差的股票每年收益率多出10%。AQR公司已设立了好几项资金以探索动量市场不规则表现。

Too costly, too risky?

成本高风险大吗？

Even the high priests of efficient-market theory have acknowledged the momentum effect. Well-paid fund managers have spent decades trying to find ways to beat the market. But you have to wonder why they bother devoting so much money and effort to researching the fortunes of individual companies when the momentum approach appears to be easy to exploit and has been around for a long time.

即便是有效市场理论的学术领导者，也已经承认了动量效应的存在。几十年来，拿高薪的基金经理人一直试图找到跑赢大盘的方法。你一定很奇怪，既然动量效应看起来这么容易利用，而且长期存在，为什么他们还把那么多金钱和精力花在对个别公司的研究上呢。

Logic suggests that the effect should be arbitraged away. If the best performers of the past 12 months continue to do well, smart investors will buy them after 11 months have elapsed, reducing the returns on offer to those who wait the extra month. In turn, others will buy after ten months, then nine, eight and so on until the effect disappears.

但是按逻辑推理，动量效应会被人们的套购行为消磨掉。如果在过去12个月中业绩最好的股票持续其良好表现，精明的投资者会在第11个月过去时买入这些股票，从而使等待12月方才买进它们的投资者获利率降低。于是，其他人就会在第10个月过去时买入这类股票，一步一步地，会有人在第9个月时买入，第8个月时买入，直到动量效应一开始就被套购行为抵消为止。

When efficient-market theorists come across a market anomaly, they tend to dismiss it in one of three ways. The first argument is that the anomaly is a statistical quirk obtained by torturing the data; it will not persist. But the momentum effect was noticed in 1985 (by Werner de Bondt, a Belgian economist now at DePaul University in Chicago, and Richard Thaler, of the University of Chicago School of Business) and has not gone away.

当有效市场理论家看到一个市场不规则表现，他们会用三种方法把这种现象解释掉。一种论证会说这样的异常现象是因为对统计数据扭曲解读造成的数字上的假象。但是动量效应在1985年就已被人们注意到了（由一位比利时经济学家，现工作于芝加哥DePaul 大学的Werner de Bondt，和芝加哥大学Booth商学院的 Richard Thaler发现），而此效应长存至今。

The second is that any gains from the strategy will be dissipated in higher trading costs. Clearly, the LBS team’s strategy of rebalancing a portfolio every month would be expensive but Mr Marsh says these would not offset an annual performance gap of over ten percentage points.

另一观点是用这种策略获得的收益会被更高的交易成本分散掉。很清楚的是，伦敦商学院研究小组的投资策略中包括每个月重新调整股票配置，这会使交易成本更高，但Marsh说较高的交易成本没有影响到这种策略每年十个百分点的收益差额。

The third is that higher returns simply reflect the higher risks of the strategy. This has been used to explain away two other notable anomalies: the size and value effects. Small companies tend to do better than bigger ones in the long term, but they tend to be less diversified and therefore more risky. And shares that look cheap on conventional measures (asset value, dividend yield, price-earnings ratio) also tend to deliver above-average returns, but belong to firms that are likelier to go bust.

第三种反对观点是这种策略的高收益恰恰体现出它的高风险。这种观点曾被用来解释掉了另外两种显著的不规则市场现象：规模效应和价值效应。小公司的长期投资收益好于大公司，但由于他们的投资不够多样化，所以包含更多风险。以常规标准看（包括资产价格，股票分红和市盈率）便宜的股票，也更容易产生高出平均水平的收益，但公司的基本面要有支撑。

According to a paper by Cliff Asness, who co-founded AQR, the better performance of momentum stocks is not merely a reflection of higher risk. He finds that the momentum effect persisted even when the data were controlled for company size and value (defined as price-to-book) criteria. Another explanation is needed.

然而AQR公司的联合创始人Cliff Asness则说，因动量效应而表现好的股票并不意味着有更高的风险。他发现即便数据受公司规模和公司价值（即市净率）的控制，动量效应仍然可见。这里需要的是另一种解释。

One possibility relates to timing. The efficient-market hypothesis assumes that new developments are instantly assimilated into asset prices. However, investors may be slow to adjust their opinions to fresh information. If they view a company unfavourably, they may dismiss an improvement in quarterly profits as a blip, rather than a change in trend. So momentum may simply represent the lag between beliefs and the new reality.

动量效应的原因之一可能是时间差。有效市场理论假设，股价的新变化会即时地反映到资产价格中。而投资者可能在接受新消息时不够快，如果他们不看好一家公司，他们可能将这家公司季报中的收益改善视为偶然，并不看做是该公司业绩发展趋势的变化。所以动量效应可能表现的是投资者旧有观念和公司新发展之间的时间断层。

Once a trend is established, a share may benefit from a bandwagon effect. Professional fund managers have to prepare regular reports for clients on the progress of their portfolios. They will naturally want to demonstrate their skills by owning shares that have been rising in price and selling those that have been falling. This “window-dressing” may add to momentum. Paul Woolley of the London School of Economics has suggested that momentum might result from an agency problem. Investors reward fund managers who have recently beaten the market; such fund managers will inevitably own the most popular shares. As they get more money from clients, such managers will put more money into their favoured stocks, giving momentum an extra boost.

另外，当一支股票的上升趋势确立以后，接下去其股价走势可能受益于从众效应。职业基金经理人必须定期向客户呈交他们的投资进展报告。为了展现投资技巧，他们自然会持有价格在上涨的股票，卖出价格下跌的股票。这种装门面的行为可能也推动了势头效应。伦敦经济学院的Paul Woolley就曾提出，动量效应是由投资机构造成的。投资者对最近跑赢大盘的基金经理会给予奖励，而这样的基金经理持有的必然是最热门的股票。当投资者投给他们更多钱时，经理们会把这些钱投到他们最看好的股票上，于是这些热门股票原有的势头就又被加了一把力道。

It is hardly a surprise that the momentum effect has been exploited by some professionals for decades. Commodity trading advisers (CTAs), also known as managed futures funds, exist to exploit the phenomenon. They take advantage of trends across a wide range of asset classes, including equities and currencies as well as raw materials. Martin Lueck was one of the three founders of AHL, one of the more successful CTAs, and now works for another trend-follower, Aspect Capital. “Trends occur because there is a disequilibrium between supply and demand,” he says. “The asset is trying to get from equilibrium price A to equilibrium price B.”

一点不奇怪，动量效应已经被一些职业投资人利用了几十年。商品交易顾问，也叫做管理期货基金，就是靠利用这种现象生存的。他们利用多种资产的价格趋势赚钱，包括股票、货币和商品原材料等。Martin Lueck 曾是较为成功的商品交易顾问公司AHL的三个创建者之一，目前为另一家趋势投资机构Aspect Capital工作。他说：“趋势的存在是由于供求关系的不平衡造成的。一项资产总是试图从均衡价格A移动到均衡价格B。”

Many of the trend-following models were developed in the late 1970s and early 1980s. They were exploited by investors such as John Henry, best known outside the financial world for owning a baseball team, the Boston Red Sox, and a football club, Liverpool (which is on a downward trend of its own). One of the simplest was to buy an asset when the 20-day moving average of its price rose above its 200-day average. In a recent study Jo?lle Miffre and Georgios Rallis of the Cass Business School in London found 13 profitable momentum strategies in commodity markets with an average annual return of 9.4% between 1979 and 2004.

许多趋势模型是在20世纪70年代末80年代初建立的。利用了这些模型的投资者中包括有John Henry，但他在金融世界之外更为知名，因为他拥有一支棒球队——波士顿红袜队，而且他还是利物浦足球俱乐部的拥有者（这家俱乐部却是处于下降趋势的）。这类模型中最简单的方法之一是当一支股票的20日均线超过200天均线时买进这支股票。在最近的一项研究中，伦敦Cass 商学院的Jo?lle Miffre 和Georgios Rallis通过对1979-2004年间大宗商品交易市场的研究，发现了13种获利的动量战略，平均年回报率达到9.4%。

Modern CTAs like Aspect and Winton (run by David Harding, another founder of AHL) devote a lot of effort to researching new ways of exploiting momentum. That has sometimes meant trading faster and faster, with a time horizon of milliseconds rather than months. However, not all market movements are part of a trend. Some are merely random fluctuations. “As you trade faster, it is easier to get misled by the noise,” says Mr Lueck. Trend-followers can get “whipsawed” in volatile markets, buying at the top of a short-term trend and then selling at a loss shortly afterwards.

现代大宗商品交易顾问公司如Aspect 和Winton（经营者是David Harding，另一位AHL的创建者）投入了大量精力研究利用动量的新方法。有些方法意味着越来越快的交易频率，以微秒而不再以月份为时间单位。但是，并非所有市场运动都是趋势的一部分，有些仅仅是不规则的波动。Lueck说：“当交易频率更快的时候，投资者更容易被噪音误导。”在易变化的市场中，趋势投资者可能遭市场洗盘，在一个短期升势的顶部买入，不久之后被迫割肉卖出。

That may be one reason why the momentum effect has not been arbitraged away: it can go horribly wrong. Just as trees do not grow to the sky, share prices do not rise for ever. The effect tends to work for the best performers over the past 12 months, but not for those that have shone for longer periods, say three or five years.

这也许是动量效应未被套利交易消磨掉的原因之一：踏空失误可以很严重。正如树木不会顶破天空，股价也不会永远上涨。动量效应对于过去12个月中表现最好的股票是起作用的，但较长期的绩优股，比如3-5年中表现好的股票，则不在此规律之内。

The value of value

价值的价值

That may be because of another anomaly, the value effect. Investors eventually get too pessimistic about struggling firms, and price their shares too cheaply. That turns them into bargains. Broadly, whereas momentum works over the short term, value is successful over longer periods. The result can be sharp reversals in markets—and nasty surprises for momentum traders. One such turning-point occurred in 2009. Investors who used a short-term momentum strategy, buying the winners of the previous six months, would have lost 46% in the British market and 53% in America, according to the LBS team. Similarly bad years were 1975, 2000 and 2003.

这种分化也许是另一种不规则市场现象——价值效应——造成的。投资者对于长期挣扎的公司渐渐悲观，给这类股票的定价很低。这就使得它们成为有利可图的便宜货。总的来讲，动量效应在短期范围内有效，价值效应在长期意义上更成功。但有时，动量交易的结果也可能是与投资者期待完全相反的、非常糟糕的意外。根据伦敦商学院研究小组的研究结果，动量转折点在2009年出现过一次，如果采用短期动量策略的投资者买进了之前6个月的绩优股的话，英国股市会使其损失投资的46%，美国股市会损失其投资的53%。相似的糟糕年头还出现在1975，2000和2003年。

The momentum effect allows investors to get rich slowly. But many fund managers are impatient and thus use leverage (borrowed money) to enhance returns. Such an approach would lose so much money in bad years that clients might lose faith. “To exploit momentum, you need investors who understand the portfolio is going to be subject to a very high level of volatility,” says Mr Marsh of the LBS.

动量效应允许投资者慢慢变富。但许多基金经理可没有什么耐心，他们会利用资金杠杆（借钱）来扩大投资收益。这种方式在不景气的年头会损失太多钱，导致客户失去信心。因此伦敦商学院的Marsh说：“要利用动量效应，你需要投资者理解的是投资组合受制于市场的高度变化。”

Momentum is so significant in stockmarkets that academics are starting to analyse what role it plays in professional fund managers’ returns. This is all part of the long process of removing the “magic” from financial performance. In the early days of fund management, in the 19th century, there were no stockmarket indices. Fund managers could thus claim that a positive return was down to their own brilliance, rather than a general rise of the market, and clients could not tell the difference.

动量效应在股票市场中是如此显著，以至于学术研究者开始分析在职业基金经理人的投资收益中它扮演的是什么角色。长期以来人们一直试图揭示操纵金融市场表现的魔术，此项研究也是这个过程的一部分。19世纪是投资基金管理的早期岁月，当时股市中并没有股票指数。于是基金经理们可以宣称投资收益是来自于他们的智慧，而非源自整个大盘的上升，他们的客户也没法区分这两者的区别。

After the development of benchmarks like the S&P 500, clients began to demand that fund managers proved their skill by outperforming an index. Many failed; but even some who succeeded may have done so by holding concentrated portfolios of only a few stocks. Such portfolios were more risky than the overall market. So the next step was to measure the managers’ performance after adjusting for risk. Even those managers may have done well because their investment style (value, for instance) was in fashion. So academics started to allow for that, too.

在如标准普尔500指数这样的基准发展出来之后，客户开始要求基金经理跑赢指数来证明他们的投资技巧。许多基金经理没有成功，即便是少数几个成功的，也是靠集中投资于几支股票。而这样的投资组合是要比整个市场更具风险的。所以接下来，人们会在针对风险做出调整后再来衡量经理人的投资表现。即便因采取了流行的投资风格（比如价值投资）而有不错表现的经理人，也要受此考验。于是研究者对此方式也视为当然了。

In effect, the portion of the investment return that was purely a result of fund managers’ skill was being reduced at every stage. Now, says Mr Marsh, academics are looking to see whether some outperformance is really all down to momentum.

事实上，随着每个阶段的发展，纯粹由基金经理的投资技巧创造的收益比例一直在降低。而如今，如Marsh所言，研究者正在探究是否某些超过大盘指数的业绩收益靠的完全是动量效应的东风。

All this analysis matters because these factors can be replicated. These days investors can not only match a benchmark through simple index-tracking funds; they can also own portfolios that exploit the value and momentum effects without paying hefty fund-management fees. The investment-management industry may become even more commoditised.

所有这些研究之所以有意义，就在于这些因素都是可复制的。如今投资者不仅可以通过指数追踪基金简单地匹配基准；他们也可以利用价值效应和动量效应自己配置股票，省掉了高额的基金管理费用。因此投资管理行业可能要变得更加商品化了。

The momentum effect raises a further important issue. If markets are rational, as the efficient-market hypothesis assumed, then they will allocate capital to its most productive uses. But the momentum effect suggests that an irrationality might be at work; investors could be buying shares (and commodities) just because they have risen in price.

动量效应提出了一个更深层次的重要问题，如果市场是理性的，如有效市场假说假设的那样，那么市场将把资金分配给收益最大的用途。但动量效应暗示却是有一种非理性在市场中运作；投资者买进股票或大宗商品仅仅因为它们的价格涨了。

That would help explain why bubbles are created and why professional investors ended up allocating capital to dotcom companies with no earnings and business plans written on the back of a cigarette packet. Momentum can carry whole economies off track.

这也许能帮我们解释市场泡沫产生的原因，以及为什么专业投资者会把钱投给毫无盈利的互联网企业，或者仅仅看到写在烟盒背面的商业计划就肯投钱。动量效应可以将整个经济带离正轨。